=======================================================   RESTART [11-22 15:50:17]   =======================================================







=======================================================   RESTART [11-22 15:50:17]   =======================================================







=======================================================   RESTART [11-22 15:50:17]   =======================================================







=======================================================   RESTART [11-22 15:50:17]   =======================================================







=======================================================   RESTART [11-22 15:50:17]   =======================================================







=======================================================   RESTART [11-22 15:50:17]   =======================================================







=======================================================   RESTART [11-22 15:50:17]   =======================================================
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:

model.safetensors:   0%|          | 0.00/346M [00:00<?, ?B/s]
model.safetensors:   6%|6         | 21.0M/346M [00:00<00:01, 181MB/s]
model.safetensors:  15%|#5        | 52.4M/346M [00:00<00:01, 224MB/s]
model.safetensors:  24%|##4       | 83.9M/346M [00:00<00:01, 235MB/s]
model.safetensors:  33%|###3      | 115M/346M [00:00<00:00, 240MB/s] 
model.safetensors:  42%|####2     | 147M/346M [00:00<00:00, 220MB/s]
model.safetensors:  51%|#####1    | 178M/346M [00:00<00:00, 229MB/s]
model.safetensors:  61%|######    | 210M/346M [00:00<00:00, 234MB/s]
model.safetensors:  70%|######9   | 241M/346M [00:01<00:00, 235MB/s]
model.safetensors:  79%|#######8  | 273M/346M [00:01<00:00, 233MB/s]
model.safetensors:  88%|########7 | 304M/346M [00:01<00:00, 234MB/s]
model.safetensors:  97%|#########6| 336M/346M [00:01<00:00, 235MB/s]
model.safetensors: 100%|##########| 346M/346M [00:01<00:00, 232MB/s]
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241121_235152-lhd87gki
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run generous-haze-537
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/lhd87gki
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_7 0.0243 ms 100.0%
  triton_mm_11 0.0246 ms 98.6%
  triton_mm_6 0.0249 ms 97.7%
  triton_mm_12 0.0250 ms 97.1%
  triton_mm_9 0.0259 ms 93.7%
  triton_mm_8 0.0261 ms 93.0%
  triton_mm_3 0.0262 ms 92.8%
  triton_mm_2 0.0264 ms 92.1%
  triton_mm_4 0.0268 ms 90.6%
  triton_mm_15 0.0272 ms 89.3%
SingleProcess AUTOTUNE benchmarking takes 2.1293 seconds and 0.0430 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0466 ms 100.0%
  triton_mm_25 0.0477 ms 97.8%
  triton_mm_33 0.0480 ms 97.1%
  triton_mm_28 0.0491 ms 95.0%
  triton_mm_18 0.0497 ms 93.9%
  triton_mm_24 0.0500 ms 93.3%
  triton_mm_19 0.0512 ms 91.1%
  triton_mm_21 0.0515 ms 90.6%
  bias_addmm 0.0516 ms 90.3%
  triton_mm_20 0.0534 ms 87.4%
SingleProcess AUTOTUNE benchmarking takes 2.2027 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3383 ms 100.0%
  triton_mm_50 0.3478 ms 97.3%
  triton_mm_52 0.3721 ms 90.9%
  triton_mm_45 0.4002 ms 84.5%
  bias_addmm 0.4026 ms 84.0%
  triton_mm_47 0.4169 ms 81.1%
  triton_mm_43 0.4295 ms 78.8%
  triton_mm_46 0.4449 ms 76.0%
  addmm 0.4633 ms 73.0%
  triton_mm_48 0.4757 ms 71.1%
SingleProcess AUTOTUNE benchmarking takes 2.6904 seconds and 0.0377 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1156 ms 100.0%
  triton_mm_69 0.1189 ms 97.3%
  triton_mm_71 0.1259 ms 91.9%
  bias_addmm 0.1287 ms 89.8%
  triton_mm_64 0.1388 ms 83.3%
  triton_mm_62 0.1446 ms 80.0%
  triton_mm_66 0.1448 ms 79.9%
  addmm 0.1464 ms 79.0%
  triton_mm_65 0.1559 ms 74.2%
  triton_mm_67 0.1663 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.5507 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4344 ms 100.0%
  triton_mm_88 0.4412 ms 98.5%
  triton_mm_90 0.4784 ms 90.8%
  bias_addmm 0.5183 ms 83.8%
  triton_mm_83 0.5324 ms 81.6%
  triton_mm_85 0.5554 ms 78.2%
  triton_mm_81 0.5626 ms 77.2%
  triton_mm_84 0.5889 ms 73.8%
  triton_mm_86 0.6270 ms 69.3%
  triton_mm_82 0.6436 ms 67.5%
SingleProcess AUTOTUNE benchmarking takes 2.7032 seconds and 0.5090 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4299 ms 100.0%
  triton_mm_108 0.4310 ms 99.7%
  triton_mm_107 0.4518 ms 95.2%
  triton_mm_109 0.4546 ms 94.6%
  bias_addmm 0.4954 ms 86.8%
  triton_mm_102 0.5364 ms 80.1%
  triton_mm_100 0.5583 ms 77.0%
  triton_mm_103 0.5602 ms 76.7%
  triton_mm_104 0.5606 ms 76.7%
  triton_mm_105 0.6271 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.6941 seconds and 0.1956 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4305 ms 100.0%
  triton_mm_200 0.4512 ms 95.4%
  triton_mm_202 0.4541 ms 94.8%
  addmm 0.4928 ms 87.3%
  bias_addmm 0.4953 ms 86.9%
  triton_mm_195 0.5363 ms 80.3%
  triton_mm_196 0.5606 ms 76.8%
  triton_mm_193 0.5607 ms 76.8%
  triton_mm_197 0.5610 ms 76.7%
  triton_mm_198 0.6253 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7007 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_294 0.4308 ms 100.0%
  triton_mm_293 0.4523 ms 95.2%
  triton_mm_295 0.4549 ms 94.7%
  addmm 0.4870 ms 88.5%
  bias_addmm 0.4945 ms 87.1%
  triton_mm_288 0.5367 ms 80.3%
  triton_mm_286 0.5596 ms 77.0%
  triton_mm_289 0.5601 ms 76.9%
  triton_mm_290 0.5623 ms 76.6%
  triton_mm_291 0.6257 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.6985 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_387 0.4298 ms 100.0%
  triton_mm_386 0.4499 ms 95.5%
  triton_mm_388 0.4554 ms 94.4%
  addmm 0.4570 ms 94.0%
  bias_addmm 0.4947 ms 86.9%
  triton_mm_381 0.5353 ms 80.3%
  triton_mm_382 0.5580 ms 77.0%
  triton_mm_383 0.5601 ms 76.7%
  triton_mm_379 0.5603 ms 76.7%
  triton_mm_384 0.6246 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7191 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_480 0.4323 ms 100.0%
  triton_mm_479 0.4507 ms 95.9%
  triton_mm_481 0.4546 ms 95.1%
  addmm 0.4925 ms 87.8%
  bias_addmm 0.4958 ms 87.2%
  triton_mm_474 0.5367 ms 80.5%
  triton_mm_472 0.5580 ms 77.5%
  triton_mm_475 0.5599 ms 77.2%
  triton_mm_476 0.5613 ms 77.0%
  triton_mm_477 0.6272 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7000 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_573 0.4299 ms 100.0%
  triton_mm_572 0.4510 ms 95.3%
  triton_mm_574 0.4537 ms 94.8%
  addmm 0.4929 ms 87.2%
  bias_addmm 0.4959 ms 86.7%
  triton_mm_567 0.5356 ms 80.3%
  triton_mm_568 0.5585 ms 77.0%
  triton_mm_569 0.5600 ms 76.8%
  triton_mm_565 0.5603 ms 76.7%
  triton_mm_570 0.6260 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7111 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1172 ms 100.0%
  triton_mm_591 0.1211 ms 96.7%
  triton_mm_593 0.1275 ms 91.9%
  triton_mm_586 0.1428 ms 82.0%
  triton_mm_584 0.1483 ms 79.0%
  triton_mm_588 0.1486 ms 78.8%
  bias_addmm 0.1543 ms 75.9%
  triton_mm_587 0.1634 ms 71.7%
  triton_mm_589 0.1670 ms 70.1%
  triton_mm_585 0.1694 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.5443 seconds and 0.0018 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1237 ms 100.0%
  triton_mm_610 0.1262 ms 98.0%
  triton_mm_612 0.1345 ms 91.9%
  bias_addmm 0.1350 ms 91.6%
  triton_mm_607 0.1524 ms 81.2%
  triton_mm_605 0.1552 ms 79.7%
  addmm 0.1554 ms 79.6%
  triton_mm_603 0.1562 ms 79.2%
  triton_mm_606 0.1739 ms 71.1%
  triton_mm_608 0.1764 ms 70.1%
SingleProcess AUTOTUNE benchmarking takes 2.5684 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0598 ms 100.0%
  triton_mm_624 0.0598 ms 99.9%
  triton_mm_626 0.0606 ms 98.6%
  triton_mm_622 0.0612 ms 97.7%
  triton_mm_630 0.0630 ms 94.9%
  triton_mm_629 0.0633 ms 94.5%
  triton_mm_631 0.0678 ms 88.1%
  triton_mm_625 0.0688 ms 86.8%
  triton_mm_627 0.0716 ms 83.5%
  triton_mm_623 0.0735 ms 81.3%
SingleProcess AUTOTUNE benchmarking takes 10.0685 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_723 0.4316 ms 100.0%
  triton_mm_722 0.4504 ms 95.8%
  triton_mm_724 0.4542 ms 95.0%
  addmm 0.4934 ms 87.5%
  bias_addmm 0.4950 ms 87.2%
  triton_mm_717 0.5370 ms 80.4%
  triton_mm_718 0.5588 ms 77.2%
  triton_mm_715 0.5596 ms 77.1%
  triton_mm_719 0.5607 ms 77.0%
  triton_mm_720 0.6267 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7137 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_816 0.4303 ms 100.0%
  triton_mm_815 0.4500 ms 95.6%
  triton_mm_817 0.4556 ms 94.5%
  addmm 0.4938 ms 87.1%
  bias_addmm 0.4953 ms 86.9%
  triton_mm_810 0.5365 ms 80.2%
  triton_mm_811 0.5585 ms 77.0%
  triton_mm_812 0.5588 ms 77.0%
  triton_mm_808 0.5607 ms 76.8%
  triton_mm_813 0.6265 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7089 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_909 0.4303 ms 100.0%
  addmm 0.4306 ms 99.9%
  triton_mm_908 0.4501 ms 95.6%
  triton_mm_910 0.4555 ms 94.5%
  bias_addmm 0.4962 ms 86.7%
  triton_mm_903 0.5351 ms 80.4%
  triton_mm_904 0.5585 ms 77.0%
  triton_mm_905 0.5605 ms 76.8%
  triton_mm_901 0.5608 ms 76.7%
  triton_mm_906 0.6267 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.6991 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_1002 0.4303 ms 100.0%
  triton_mm_1001 0.4500 ms 95.6%
  addmm 0.4529 ms 95.0%
  triton_mm_1003 0.4551 ms 94.5%
  bias_addmm 0.4951 ms 86.9%
  triton_mm_996 0.5357 ms 80.3%
  triton_mm_997 0.5587 ms 77.0%
  triton_mm_998 0.5603 ms 76.8%
  triton_mm_994 0.5604 ms 76.8%
  triton_mm_999 0.6268 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.7129 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2872 0.0228 ms 100.0%
  triton_mm_2876 0.0230 ms 99.3%
  triton_mm_2868 0.0239 ms 95.7%
  triton_mm_2880 0.0242 ms 94.6%
  bias_addmm 0.0267 ms 85.6%
  triton_mm_2865 0.0270 ms 84.7%
  triton_mm_2875 0.0276 ms 82.7%
  triton_mm_2867 0.0276 ms 82.6%
  triton_mm_2871 0.0277 ms 82.5%
  triton_mm_2866 0.0292 ms 78.3%
SingleProcess AUTOTUNE benchmarking takes 8.0055 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  bias_addmm 1.8206 ms 100.0%
  triton_mm_2898 1.8571 ms 98.0%
  triton_mm_2897 1.8805 ms 96.8%
  triton_mm_2899 2.0434 ms 89.1%
  addmm 2.1758 ms 83.7%
  triton_mm_2892 2.2761 ms 80.0%
  triton_mm_2894 2.3413 ms 77.8%
  triton_mm_2890 2.3985 ms 75.9%
  triton_mm_2893 2.4670 ms 73.8%
  triton_mm_2895 2.6952 ms 67.5%
SingleProcess AUTOTUNE benchmarking takes 11.6688 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  mm 1.7017 ms 100.0%
  triton_mm_2935 1.7053 ms 99.8%
  triton_mm_2936 1.7761 ms 95.8%
  triton_mm_2937 2.0802 ms 81.8%
  triton_mm_2928 2.2315 ms 76.3%
  triton_mm_2930 2.2624 ms 75.2%
  triton_mm_2932 2.2909 ms 74.3%
  triton_mm_2931 2.3792 ms 71.5%
  triton_mm_2933 2.4386 ms 69.8%
  triton_mm_2929 2.4701 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 9.9082 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.6953 ms 100.0%
  triton_mm_2916 1.8474 ms 91.8%
  triton_mm_2918 1.8806 ms 90.1%
  triton_mm_2917 1.8944 ms 89.5%
  triton_mm_2912 2.3383 ms 72.5%
  triton_mm_2911 2.4321 ms 69.7%
  triton_mm_2909 2.4779 ms 68.4%
  triton_mm_2913 2.4877 ms 68.1%
  triton_mm_2910 2.5532 ms 66.4%
  triton_mm_2914 2.6343 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 10.0563 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4225 ms 100.0%
  mm 0.4270 ms 98.9%
  triton_mm_3007 0.4345 ms 97.2%
  triton_mm_3008 0.4598 ms 91.9%
  triton_mm_3003 0.5231 ms 80.8%
  triton_mm_2999 0.5329 ms 79.3%
  triton_mm_3001 0.5383 ms 78.5%
  triton_mm_3004 0.5648 ms 74.8%
  triton_mm_3002 0.5700 ms 74.1%
  triton_mm_3000 0.5796 ms 72.9%
SingleProcess AUTOTUNE benchmarking takes 9.8033 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2962 0.0163 ms 100.0%
  triton_mm_2960 0.0164 ms 99.8%
  triton_mm_2967 0.0166 ms 98.3%
  triton_mm_2963 0.0169 ms 96.6%
  triton_mm_2965 0.0169 ms 96.6%
  triton_mm_2966 0.0176 ms 92.7%
  triton_mm_2964 0.0178 ms 91.7%
  triton_mm_2969 0.0179 ms 91.1%
  triton_mm_2961 0.0181 ms 90.2%
  triton_mm_2957 0.0185 ms 88.2%
SingleProcess AUTOTUNE benchmarking takes 5.3226 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0218 ms 100.0%
  triton_mm_2942 0.0253 ms 85.9%
  triton_mm_2950 0.0259 ms 83.9%
  mm 0.0284 ms 76.7%
  triton_mm_2954 0.0298 ms 73.0%
  triton_mm_2941 0.0347 ms 62.7%
  triton_mm_2940 0.0351 ms 61.9%
  triton_mm_2945 0.0375 ms 58.0%
  triton_mm_2949 0.0390 ms 55.8%
  triton_mm_2939 0.0408 ms 53.3%
SingleProcess AUTOTUNE benchmarking takes 6.5783 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4154 ms 100.0%
  triton_mm_2988 0.4204 ms 98.8%
  mm 0.4407 ms 94.3%
  triton_mm_2989 0.4558 ms 91.1%
  triton_mm_2982 0.5197 ms 79.9%
  triton_mm_2980 0.5303 ms 78.3%
  triton_mm_2984 0.5314 ms 78.2%
  triton_mm_2983 0.5690 ms 73.0%
  triton_mm_2981 0.5893 ms 70.5%
  triton_mm_2985 0.6028 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 10.1873 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4163 ms 100.0%
  triton_mm_3044 0.4199 ms 99.1%
  triton_mm_3045 0.4369 ms 95.3%
  triton_mm_3046 0.4622 ms 90.1%
  triton_mm_3037 0.5184 ms 80.3%
  triton_mm_3041 0.5269 ms 79.0%
  triton_mm_3039 0.5278 ms 78.9%
  triton_mm_3040 0.5556 ms 74.9%
  triton_mm_3042 0.5686 ms 73.2%
  triton_mm_3038 0.5730 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 9.8798 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4012 ms 100.0%
  triton_mm_3025 0.4286 ms 93.6%
  triton_mm_3026 0.4404 ms 91.1%
  triton_mm_3027 0.4448 ms 90.2%
  triton_mm_3020 0.5206 ms 77.1%
  triton_mm_3018 0.5330 ms 75.3%
  triton_mm_3022 0.5389 ms 74.5%
  triton_mm_3021 0.5438 ms 73.8%
  triton_mm_3019 0.5932 ms 67.6%
  triton_mm_3023 0.6174 ms 65.0%
SingleProcess AUTOTUNE benchmarking takes 9.7832 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3204 ms 100.0%
  triton_mm_3120 0.3243 ms 98.8%
  triton_mm_3121 0.3379 ms 94.8%
  triton_mm_3122 0.3559 ms 90.0%
  triton_mm_3113 0.3978 ms 80.5%
  triton_mm_3115 0.3990 ms 80.3%
  triton_mm_3117 0.4018 ms 79.7%
  triton_mm_3116 0.4237 ms 75.6%
  triton_mm_3118 0.4327 ms 74.0%
  triton_mm_3114 0.4366 ms 73.4%
SingleProcess AUTOTUNE benchmarking takes 9.9503 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3204 ms 100.0%
  triton_mm_3102 0.3224 ms 99.4%
  mm 0.3305 ms 97.0%
  triton_mm_3103 0.3358 ms 95.4%
  triton_mm_3096 0.3892 ms 82.3%
  triton_mm_3094 0.4035 ms 79.4%
  triton_mm_3098 0.4061 ms 78.9%
  triton_mm_3097 0.4071 ms 78.7%
  triton_mm_3095 0.4455 ms 71.9%
  triton_mm_3099 0.4620 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 9.8980 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1403 ms 100.0%
  triton_mm_3075 0.1463 ms 95.8%
  triton_mm_3083 0.1476 ms 95.1%
  triton_mm_3079 0.1478 ms 94.9%
  triton_mm_3084 0.1501 ms 93.4%
  triton_mm_3080 0.1674 ms 83.8%
  triton_mm_3076 0.1695 ms 82.7%
  mm 0.1769 ms 79.3%
  triton_mm_3077 0.1780 ms 78.8%
  triton_mm_3074 0.1826 ms 76.8%
SingleProcess AUTOTUNE benchmarking takes 13.7807 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1104 ms 100.0%
  triton_mm_3063 0.1113 ms 99.2%
  triton_mm_3064 0.1140 ms 96.8%
  triton_mm_3065 0.1219 ms 90.5%
  triton_mm_3058 0.1331 ms 82.9%
  triton_mm_3056 0.1388 ms 79.5%
  triton_mm_3060 0.1413 ms 78.1%
  triton_mm_3059 0.1495 ms 73.8%
  triton_mm_3057 0.1557 ms 70.9%
  triton_mm_3061 0.1613 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 10.2749 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3149 0.0327 ms 100.0%
  triton_mm_3150 0.0337 ms 97.1%
  triton_mm_3152 0.0337 ms 97.0%
  triton_mm_3154 0.0338 ms 96.6%
  triton_mm_3147 0.0340 ms 96.2%
  triton_mm_3145 0.0340 ms 96.1%
  triton_mm_3146 0.0341 ms 95.8%
  triton_mm_3151 0.0341 ms 95.8%
  triton_mm_3148 0.0347 ms 94.2%
  triton_mm_3153 0.0354 ms 92.3%
SingleProcess AUTOTUNE benchmarking takes 5.3458 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4250 ms 100.0%
  mm 0.4272 ms 99.5%
  triton_mm_3192 0.4391 ms 96.8%
  triton_mm_3193 0.4657 ms 91.3%
  triton_mm_3188 0.5307 ms 80.1%
  triton_mm_3184 0.5388 ms 78.9%
  triton_mm_3186 0.5461 ms 77.8%
  triton_mm_3189 0.5745 ms 74.0%
  triton_mm_3187 0.5781 ms 73.5%
  triton_mm_3185 0.5867 ms 72.4%
SingleProcess AUTOTUNE benchmarking takes 2.5699 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0531 ms 100.0%
  triton_mm_3131 0.0533 ms 99.6%
  triton_mm_3127 0.0537 ms 98.9%
  triton_mm_3135 0.0588 ms 90.3%
  triton_mm_3139 0.0683 ms 77.8%
  triton_mm_3126 0.0846 ms 62.8%
  triton_mm_3125 0.0879 ms 60.4%
  triton_mm_3130 0.0956 ms 55.6%
  triton_mm_3134 0.1013 ms 52.4%
  triton_mm_3124 0.1028 ms 51.6%
SingleProcess AUTOTUNE benchmarking takes 6.5445 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4185 ms 100.0%
  triton_mm_3173 0.4272 ms 98.0%
  mm 0.4391 ms 95.3%
  triton_mm_3174 0.4632 ms 90.4%
  triton_mm_3167 0.5260 ms 79.6%
  triton_mm_3165 0.5373 ms 77.9%
  triton_mm_3169 0.5375 ms 77.9%
  triton_mm_3168 0.5764 ms 72.6%
  triton_mm_3166 0.5961 ms 70.2%
  triton_mm_3170 0.6152 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.5755 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4151 ms 100.0%
  triton_mm_3229 0.4243 ms 97.8%
  triton_mm_3230 0.4420 ms 93.9%
  triton_mm_3231 0.4693 ms 88.4%
  triton_mm_3222 0.5262 ms 78.9%
  triton_mm_3226 0.5330 ms 77.9%
  triton_mm_3224 0.5367 ms 77.3%
  triton_mm_3225 0.5647 ms 73.5%
  triton_mm_3227 0.5770 ms 71.9%
  triton_mm_3223 0.5829 ms 71.2%
SingleProcess AUTOTUNE benchmarking takes 2.5654 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4022 ms 100.0%
  triton_mm_3210 0.4320 ms 93.1%
  triton_mm_3211 0.4470 ms 90.0%
  triton_mm_3212 0.4521 ms 89.0%
  triton_mm_3205 0.5278 ms 76.2%
  triton_mm_3203 0.5421 ms 74.2%
  triton_mm_3207 0.5470 ms 73.5%
  triton_mm_3206 0.5520 ms 72.9%
  triton_mm_3204 0.6030 ms 66.7%
  triton_mm_3208 0.6254 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.5812 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  triton_mm_3305 0.3271 ms 100.0%
  triton_mm_3306 0.3403 ms 96.1%
  triton_mm_3307 0.3606 ms 90.7%
  mm 0.3694 ms 88.5%
  triton_mm_3298 0.4010 ms 81.6%
  triton_mm_3300 0.4047 ms 80.8%
  triton_mm_3302 0.4056 ms 80.6%
  triton_mm_3301 0.4284 ms 76.4%
  triton_mm_3303 0.4377 ms 74.7%
  triton_mm_3299 0.4408 ms 74.2%
SingleProcess AUTOTUNE benchmarking takes 2.5450 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3235 ms 100.0%
  triton_mm_3287 0.3267 ms 99.0%
  mm 0.3313 ms 97.7%
  triton_mm_3288 0.3397 ms 95.2%
  triton_mm_3281 0.3943 ms 82.1%
  triton_mm_3279 0.4084 ms 79.2%
  triton_mm_3283 0.4116 ms 78.6%
  triton_mm_3282 0.4126 ms 78.4%
  triton_mm_3280 0.4503 ms 71.8%
  triton_mm_3284 0.4675 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.5530 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1415 ms 100.0%
  triton_mm_3260 0.1462 ms 96.8%
  triton_mm_3268 0.1481 ms 95.6%
  triton_mm_3264 0.1482 ms 95.5%
  triton_mm_3262 0.1488 ms 95.1%
  triton_mm_3269 0.1500 ms 94.4%
  triton_mm_3263 0.1644 ms 86.0%
  triton_mm_3265 0.1690 ms 83.7%
  triton_mm_3261 0.1701 ms 83.2%
  mm 0.1799 ms 78.6%
SingleProcess AUTOTUNE benchmarking takes 2.4527 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1108 ms 100.0%
  triton_mm_3248 0.1128 ms 98.3%
  triton_mm_3249 0.1144 ms 96.9%
  triton_mm_3250 0.1216 ms 91.1%
  triton_mm_3243 0.1338 ms 82.9%
  triton_mm_3241 0.1394 ms 79.5%
  triton_mm_3245 0.1414 ms 78.4%
  triton_mm_3244 0.1495 ms 74.1%
  triton_mm_3242 0.1556 ms 71.3%
  triton_mm_3246 0.1620 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.4471 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3334 0.0327 ms 100.0%
  triton_mm_3336 0.0330 ms 99.1%
  triton_mm_3337 0.0338 ms 96.8%
  triton_mm_3330 0.0340 ms 96.1%
  triton_mm_3332 0.0340 ms 96.1%
  triton_mm_3331 0.0341 ms 96.0%
  triton_mm_3335 0.0346 ms 94.5%
  triton_mm_3333 0.0347 ms 94.3%
  triton_mm_3339 0.0349 ms 93.6%
  triton_mm_3338 0.0366 ms 89.3%
SingleProcess AUTOTUNE benchmarking takes 2.0213 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3316 0.0528 ms 100.0%
  mm 0.0530 ms 99.7%
  triton_mm_3312 0.0532 ms 99.3%
  triton_mm_3320 0.0589 ms 89.7%
  triton_mm_3324 0.0695 ms 76.0%
  triton_mm_3311 0.0838 ms 63.0%
  triton_mm_3310 0.0887 ms 59.5%
  triton_mm_3315 0.0948 ms 55.7%
  triton_mm_3319 0.1012 ms 52.1%
  triton_mm_3309 0.1036 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 2.1522 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3357 0.4167 ms 100.0%
  triton_mm_3358 0.4234 ms 98.4%
  mm 0.4378 ms 95.2%
  triton_mm_3359 0.4627 ms 90.1%
  triton_mm_3352 0.5250 ms 79.4%
  triton_mm_3350 0.5354 ms 77.8%
  triton_mm_3354 0.5365 ms 77.7%
  triton_mm_3353 0.5760 ms 72.3%
  triton_mm_3351 0.6011 ms 69.3%
  triton_mm_3355 0.6113 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 2.5768 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4190 ms 100.0%
  triton_mm_3414 0.4238 ms 98.9%
  triton_mm_3415 0.4404 ms 95.1%
  triton_mm_3416 0.4667 ms 89.8%
  triton_mm_3407 0.5244 ms 79.9%
  triton_mm_3409 0.5325 ms 78.7%
  triton_mm_3411 0.5335 ms 78.5%
  triton_mm_3410 0.5604 ms 74.8%
  triton_mm_3412 0.5762 ms 72.7%
  triton_mm_3408 0.5795 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 2.5727 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4013 ms 100.0%
  triton_mm_3395 0.4312 ms 93.1%
  triton_mm_3396 0.4448 ms 90.2%
  triton_mm_3397 0.4497 ms 89.2%
  triton_mm_3390 0.5276 ms 76.1%
  triton_mm_3388 0.5413 ms 74.1%
  triton_mm_3392 0.5453 ms 73.6%
  triton_mm_3391 0.5517 ms 72.7%
  triton_mm_3389 0.5993 ms 67.0%
  triton_mm_3393 0.6213 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.5861 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3225 ms 100.0%
  triton_mm_3490 0.3260 ms 98.9%
  triton_mm_3491 0.3410 ms 94.6%
  triton_mm_3492 0.3597 ms 89.7%
  triton_mm_3483 0.4008 ms 80.5%
  triton_mm_3485 0.4034 ms 80.0%
  triton_mm_3487 0.4055 ms 79.5%
  triton_mm_3486 0.4284 ms 75.3%
  triton_mm_3488 0.4370 ms 73.8%
  triton_mm_3484 0.4400 ms 73.3%
SingleProcess AUTOTUNE benchmarking takes 2.5524 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3471 0.3229 ms 100.0%
  triton_mm_3472 0.3243 ms 99.6%
  mm 0.3319 ms 97.3%
  triton_mm_3473 0.3401 ms 94.9%
  triton_mm_3466 0.3939 ms 82.0%
  triton_mm_3464 0.4064 ms 79.4%
  triton_mm_3468 0.4118 ms 78.4%
  triton_mm_3467 0.4122 ms 78.3%
  triton_mm_3465 0.4508 ms 71.6%
  triton_mm_3469 0.4667 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.5567 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3452 0.1405 ms 100.0%
  triton_mm_3445 0.1468 ms 95.7%
  triton_mm_3449 0.1481 ms 94.9%
  triton_mm_3453 0.1487 ms 94.5%
  triton_mm_3447 0.1487 ms 94.5%
  triton_mm_3454 0.1507 ms 93.3%
  triton_mm_3448 0.1643 ms 85.5%
  triton_mm_3450 0.1678 ms 83.8%
  triton_mm_3446 0.1706 ms 82.4%
  mm 0.1794 ms 78.3%
SingleProcess AUTOTUNE benchmarking takes 2.4561 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1115 ms 100.0%
  triton_mm_3433 0.1127 ms 99.0%
  triton_mm_3434 0.1143 ms 97.6%
  triton_mm_3435 0.1214 ms 91.8%
  triton_mm_3428 0.1337 ms 83.4%
  triton_mm_3426 0.1391 ms 80.2%
  triton_mm_3430 0.1420 ms 78.5%
  triton_mm_3429 0.1489 ms 74.9%
  triton_mm_3427 0.1553 ms 71.8%
  triton_mm_3431 0.1619 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.4536 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3637 0.1404 ms 100.0%
  triton_mm_3630 0.1456 ms 96.4%
  triton_mm_3638 0.1476 ms 95.1%
  triton_mm_3634 0.1479 ms 94.9%
  triton_mm_3632 0.1482 ms 94.7%
  triton_mm_3639 0.1503 ms 93.4%
  triton_mm_3633 0.1645 ms 85.3%
  triton_mm_3635 0.1676 ms 83.7%
  triton_mm_3631 0.1695 ms 82.8%
  mm 0.1771 ms 79.3%
SingleProcess AUTOTUNE benchmarking takes 2.4564 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1110 ms 100.0%
  triton_mm_3618 0.1125 ms 98.7%
  triton_mm_3619 0.1141 ms 97.3%
  triton_mm_3620 0.1212 ms 91.6%
  triton_mm_3613 0.1332 ms 83.3%
  triton_mm_3611 0.1380 ms 80.4%
  triton_mm_3615 0.1412 ms 78.6%
  triton_mm_3614 0.1490 ms 74.5%
  triton_mm_3612 0.1550 ms 71.6%
  triton_mm_3616 0.1616 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.4519 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0556 ms 100.0%
  mm 0.0567 ms 98.1%
  triton_mm_7428 0.0584 ms 95.2%
  triton_mm_7420 0.0642 ms 86.5%
  triton_mm_7424 0.0664 ms 83.8%
  triton_mm_7422 0.0676 ms 82.2%
  triton_mm_7421 0.0700 ms 79.5%
  triton_mm_7429 0.0723 ms 76.9%
  triton_mm_7425 0.0738 ms 75.3%
  triton_mm_7423 0.0855 ms 65.0%
SingleProcess AUTOTUNE benchmarking takes 10.5782 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1180 ms 100.0%
  triton_mm_7465 0.1218 ms 96.9%
  triton_mm_7466 0.1259 ms 93.8%
  triton_mm_7467 0.1323 ms 89.2%
  triton_mm_7458 0.1536 ms 76.8%
  triton_mm_7460 0.1541 ms 76.6%
  triton_mm_7462 0.1565 ms 75.4%
  triton_mm_7459 0.1667 ms 70.8%
  triton_mm_7461 0.1683 ms 70.1%
  triton_mm_7463 0.1699 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 10.2473 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1269 ms 100.0%
  triton_mm_7484 0.1444 ms 87.9%
  triton_mm_7485 0.1503 ms 84.5%
  triton_mm_7481 0.1515 ms 83.8%
  triton_mm_7486 0.1533 ms 82.8%
  triton_mm_7477 0.1552 ms 81.8%
  triton_mm_7479 0.1573 ms 80.7%
  triton_mm_7480 0.1681 ms 75.5%
  triton_mm_7482 0.1707 ms 74.3%
  triton_mm_7478 0.1729 ms 73.4%
SingleProcess AUTOTUNE benchmarking takes 9.6446 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1303 ms 100.0%
  triton_mm_7522 0.1429 ms 91.2%
  triton_mm_7523 0.1491 ms 87.4%
  triton_mm_7515 0.1495 ms 87.1%
  triton_mm_7519 0.1502 ms 86.7%
  triton_mm_7524 0.1530 ms 85.1%
  triton_mm_7517 0.1547 ms 84.2%
  triton_mm_7520 0.1689 ms 77.1%
  triton_mm_7516 0.1708 ms 76.3%
  triton_mm_7514 0.1928 ms 67.6%
SingleProcess AUTOTUNE benchmarking takes 9.7825 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0569 ms 100.0%
  triton_mm_7447 0.0614 ms 92.6%
  triton_mm_7441 0.0638 ms 89.1%
  triton_mm_7448 0.0676 ms 84.1%
  triton_mm_7443 0.0703 ms 80.8%
  triton_mm_7439 0.0711 ms 79.9%
  triton_mm_7446 0.0712 ms 79.9%
  triton_mm_7444 0.0722 ms 78.8%
  triton_mm_7442 0.0724 ms 78.5%
  triton_mm_7440 0.0748 ms 76.0%
SingleProcess AUTOTUNE benchmarking takes 9.8814 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1183 ms 100.0%
  triton_mm_7503 0.1200 ms 98.6%
  triton_mm_7504 0.1254 ms 94.3%
  triton_mm_7505 0.1282 ms 92.3%
  triton_mm_7500 0.1461 ms 81.0%
  triton_mm_7498 0.1479 ms 80.0%
  triton_mm_7496 0.1479 ms 80.0%
  triton_mm_7499 0.1571 ms 75.3%
  triton_mm_7497 0.1652 ms 71.6%
  triton_mm_7501 0.1702 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 10.4107 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0321 ms 100.0%
  triton_mm_8637 0.0348 ms 92.2%
  triton_mm_8638 0.0406 ms 79.1%
  triton_mm_8636 0.0409 ms 78.6%
  triton_mm_8645 0.0410 ms 78.4%
  triton_mm_8642 0.0420 ms 76.5%
  triton_mm_8646 0.0438 ms 73.3%
  triton_mm_8639 0.0444 ms 72.3%
  triton_mm_8644 0.0462 ms 69.6%
  triton_mm_8650 0.0470 ms 68.3%
SingleProcess AUTOTUNE benchmarking takes 7.2406 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_11 0.0246 ms 100.0%
  triton_mm_8 0.0247 ms 99.6%
  triton_mm_6 0.0248 ms 99.0%
  triton_mm_12 0.0250 ms 98.5%
  triton_mm_7 0.0252 ms 97.6%
  triton_mm_9 0.0256 ms 95.9%
  triton_mm_3 0.0260 ms 94.5%
  triton_mm_4 0.0265 ms 92.6%
  triton_mm_15 0.0270 ms 90.9%
  triton_mm_16 0.0284 ms 86.5%
SingleProcess AUTOTUNE benchmarking takes 2.0418 seconds and 0.0208 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0466 ms 100.0%
  triton_mm_25 0.0476 ms 98.0%
  triton_mm_33 0.0486 ms 95.9%
  triton_mm_28 0.0500 ms 93.3%
  triton_mm_24 0.0500 ms 93.2%
  triton_mm_18 0.0502 ms 92.9%
  triton_mm_21 0.0506 ms 92.2%
  bias_addmm 0.0516 ms 90.3%
  triton_mm_19 0.0517 ms 90.2%
  triton_mm_31 0.0539 ms 86.5%
SingleProcess AUTOTUNE benchmarking takes 2.2036 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3388 ms 100.0%
  triton_mm_50 0.3468 ms 97.7%
  triton_mm_52 0.3712 ms 91.3%
  triton_mm_45 0.3987 ms 85.0%
  bias_addmm 0.4009 ms 84.5%
  triton_mm_47 0.4159 ms 81.5%
  triton_mm_43 0.4247 ms 79.8%
  triton_mm_46 0.4448 ms 76.2%
  addmm 0.4575 ms 74.0%
  triton_mm_48 0.4760 ms 71.2%
SingleProcess AUTOTUNE benchmarking takes 2.7281 seconds and 0.0332 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1157 ms 100.0%
  triton_mm_69 0.1189 ms 97.3%
  triton_mm_71 0.1260 ms 91.8%
  bias_addmm 0.1285 ms 90.0%
  triton_mm_64 0.1383 ms 83.7%
  triton_mm_62 0.1458 ms 79.4%
  triton_mm_66 0.1459 ms 79.3%
  addmm 0.1462 ms 79.1%
  triton_mm_65 0.1559 ms 74.2%
  triton_mm_67 0.1656 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 2.5507 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4336 ms 100.0%
  triton_mm_88 0.4425 ms 98.0%
  triton_mm_90 0.4766 ms 91.0%
  bias_addmm 0.5172 ms 83.8%
  triton_mm_83 0.5300 ms 81.8%
  triton_mm_85 0.5536 ms 78.3%
  triton_mm_81 0.5642 ms 76.9%
  triton_mm_84 0.5880 ms 73.7%
  triton_mm_86 0.6308 ms 68.7%
  triton_mm_82 0.6416 ms 67.6%
SingleProcess AUTOTUNE benchmarking takes 2.7094 seconds and 1.2881 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_108 0.4297 ms 100.0%
  addmm 0.4331 ms 99.2%
  triton_mm_107 0.4508 ms 95.3%
  triton_mm_109 0.4539 ms 94.7%
  bias_addmm 0.4929 ms 87.2%
  triton_mm_102 0.5370 ms 80.0%
  triton_mm_100 0.5580 ms 77.0%
  triton_mm_104 0.5599 ms 76.8%
  triton_mm_103 0.5603 ms 76.7%
  triton_mm_105 0.6236 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7050 seconds and 0.1760 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_122 0.0466 ms 100.0%
  triton_mm_118 0.0484 ms 96.3%
  triton_mm_126 0.0487 ms 95.7%
  triton_mm_121 0.0491 ms 95.0%
  triton_mm_117 0.0500 ms 93.3%
  triton_mm_111 0.0501 ms 93.1%
  triton_mm_114 0.0507 ms 92.0%
  triton_mm_112 0.0516 ms 90.4%
  bias_addmm 0.0517 ms 90.1%
  triton_mm_124 0.0540 ms 86.3%
SingleProcess AUTOTUNE benchmarking takes 2.1904 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4315 ms 100.0%
  triton_mm_200 0.4516 ms 95.5%
  triton_mm_202 0.4550 ms 94.8%
  addmm 0.4642 ms 93.0%
  bias_addmm 0.4929 ms 87.5%
  triton_mm_195 0.5344 ms 80.7%
  triton_mm_193 0.5577 ms 77.4%
  triton_mm_196 0.5582 ms 77.3%
  triton_mm_197 0.5602 ms 77.0%
  triton_mm_198 0.6232 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7041 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_215 0.0476 ms 100.0%
  triton_mm_211 0.0484 ms 98.2%
  triton_mm_219 0.0488 ms 97.6%
  triton_mm_210 0.0494 ms 96.4%
  triton_mm_204 0.0495 ms 96.1%
  triton_mm_214 0.0502 ms 94.9%
  triton_mm_207 0.0515 ms 92.4%
  triton_mm_205 0.0518 ms 91.9%
  bias_addmm 0.0520 ms 91.6%
  triton_mm_206 0.0537 ms 88.5%
SingleProcess AUTOTUNE benchmarking takes 2.2047 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_294 0.4316 ms 100.0%
  triton_mm_293 0.4519 ms 95.5%
  triton_mm_295 0.4582 ms 94.2%
  addmm 0.4917 ms 87.8%
  bias_addmm 0.4937 ms 87.4%
  triton_mm_288 0.5319 ms 81.1%
  triton_mm_286 0.5581 ms 77.3%
  triton_mm_289 0.5586 ms 77.3%
  triton_mm_290 0.5608 ms 77.0%
  triton_mm_291 0.6236 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7034 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_308 0.0466 ms 100.0%
  triton_mm_312 0.0481 ms 96.9%
  triton_mm_304 0.0484 ms 96.2%
  triton_mm_303 0.0495 ms 94.1%
  triton_mm_297 0.0500 ms 93.1%
  triton_mm_307 0.0502 ms 92.8%
  triton_mm_300 0.0507 ms 91.9%
  triton_mm_298 0.0510 ms 91.3%
  bias_addmm 0.0520 ms 89.6%
  triton_mm_299 0.0538 ms 86.6%
SingleProcess AUTOTUNE benchmarking takes 2.2041 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_387 0.4307 ms 100.0%
  addmm 0.4333 ms 99.4%
  triton_mm_386 0.4519 ms 95.3%
  triton_mm_388 0.4544 ms 94.8%
  bias_addmm 0.4929 ms 87.4%
  triton_mm_381 0.5322 ms 80.9%
  triton_mm_382 0.5590 ms 77.0%
  triton_mm_379 0.5595 ms 77.0%
  triton_mm_383 0.5603 ms 76.9%
  triton_mm_384 0.6240 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7047 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_401 0.0466 ms 100.0%
  triton_mm_397 0.0484 ms 96.2%
  triton_mm_405 0.0488 ms 95.5%
  triton_mm_400 0.0490 ms 95.1%
  triton_mm_390 0.0494 ms 94.2%
  triton_mm_396 0.0502 ms 92.7%
  triton_mm_393 0.0506 ms 92.0%
  bias_addmm 0.0516 ms 90.2%
  triton_mm_391 0.0519 ms 89.7%
  triton_mm_403 0.0541 ms 86.1%
SingleProcess AUTOTUNE benchmarking takes 2.2226 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_480 0.4299 ms 100.0%
  addmm 0.4334 ms 99.2%
  triton_mm_479 0.4510 ms 95.3%
  triton_mm_481 0.4529 ms 94.9%
  bias_addmm 0.4944 ms 87.0%
  triton_mm_474 0.5317 ms 80.8%
  triton_mm_475 0.5580 ms 77.0%
  triton_mm_472 0.5591 ms 76.9%
  triton_mm_476 0.5608 ms 76.7%
  triton_mm_477 0.6227 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7081 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_494 0.0467 ms 100.0%
  triton_mm_490 0.0476 ms 97.9%
  triton_mm_498 0.0481 ms 97.0%
  triton_mm_493 0.0491 ms 95.1%
  triton_mm_483 0.0500 ms 93.3%
  triton_mm_489 0.0501 ms 93.1%
  triton_mm_486 0.0507 ms 92.0%
  triton_mm_484 0.0510 ms 91.4%
  bias_addmm 0.0521 ms 89.6%
  triton_mm_496 0.0541 ms 86.3%
SingleProcess AUTOTUNE benchmarking takes 2.2175 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_573 0.4297 ms 100.0%
  triton_mm_572 0.4508 ms 95.3%
  triton_mm_574 0.4534 ms 94.8%
  addmm 0.4915 ms 87.4%
  bias_addmm 0.4937 ms 87.0%
  triton_mm_567 0.5332 ms 80.6%
  triton_mm_569 0.5565 ms 77.2%
  triton_mm_565 0.5580 ms 77.0%
  triton_mm_568 0.5581 ms 77.0%
  triton_mm_570 0.6228 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7152 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1184 ms 100.0%
  triton_mm_591 0.1216 ms 97.4%
  triton_mm_593 0.1291 ms 91.7%
  triton_mm_586 0.1439 ms 82.3%
  triton_mm_584 0.1501 ms 78.9%
  triton_mm_588 0.1519 ms 77.9%
  bias_addmm 0.1540 ms 76.9%
  triton_mm_587 0.1625 ms 72.9%
  triton_mm_589 0.1665 ms 71.1%
  triton_mm_585 0.1681 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 2.5171 seconds and 0.0249 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1244 ms 100.0%
  triton_mm_610 0.1273 ms 97.7%
  triton_mm_612 0.1351 ms 92.0%
  bias_addmm 0.1364 ms 91.1%
  triton_mm_605 0.1545 ms 80.5%
  addmm 0.1549 ms 80.3%
  triton_mm_607 0.1551 ms 80.2%
  triton_mm_603 0.1569 ms 79.3%
  triton_mm_606 0.1733 ms 71.7%
  triton_mm_608 0.1756 ms 70.8%
SingleProcess AUTOTUNE benchmarking takes 11.0078 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0594 ms 100.0%
  triton_mm_624 0.0604 ms 98.4%
  triton_mm_626 0.0612 ms 97.2%
  triton_mm_622 0.0623 ms 95.4%
  triton_mm_629 0.0629 ms 94.5%
  triton_mm_630 0.0637 ms 93.2%
  triton_mm_625 0.0678 ms 87.7%
  triton_mm_631 0.0679 ms 87.5%
  triton_mm_627 0.0722 ms 82.4%
  triton_mm_623 0.0732 ms 81.2%
SingleProcess AUTOTUNE benchmarking takes 2.3617 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_644 0.0467 ms 100.0%
  triton_mm_640 0.0484 ms 96.4%
  triton_mm_648 0.0487 ms 95.8%
  triton_mm_633 0.0500 ms 93.3%
  triton_mm_639 0.0500 ms 93.3%
  triton_mm_643 0.0500 ms 93.3%
  triton_mm_634 0.0509 ms 91.6%
  triton_mm_636 0.0515 ms 90.7%
  bias_addmm 0.0515 ms 90.6%
  triton_mm_646 0.0540 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 2.2276 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_723 0.4305 ms 100.0%
  addmm 0.4345 ms 99.1%
  triton_mm_722 0.4530 ms 95.0%
  triton_mm_724 0.4539 ms 94.9%
  bias_addmm 0.4934 ms 87.2%
  triton_mm_717 0.5317 ms 81.0%
  triton_mm_718 0.5584 ms 77.1%
  triton_mm_715 0.5597 ms 76.9%
  triton_mm_719 0.5599 ms 76.9%
  triton_mm_720 0.6227 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.7117 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_816 0.4301 ms 100.0%
  triton_mm_815 0.4510 ms 95.4%
  triton_mm_817 0.4544 ms 94.6%
  addmm 0.4916 ms 87.5%
  bias_addmm 0.4938 ms 87.1%
  triton_mm_810 0.5348 ms 80.4%
  triton_mm_812 0.5570 ms 77.2%
  triton_mm_811 0.5586 ms 77.0%
  triton_mm_808 0.5590 ms 76.9%
  triton_mm_813 0.6243 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7107 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2872 0.0224 ms 100.0%
  triton_mm_2876 0.0227 ms 98.7%
  triton_mm_2868 0.0237 ms 94.6%
  triton_mm_2880 0.0239 ms 94.0%
  bias_addmm 0.0250 ms 89.6%
  triton_mm_2865 0.0269 ms 83.5%
  addmm 0.0271 ms 82.9%
  triton_mm_2871 0.0275 ms 81.6%
  triton_mm_2867 0.0275 ms 81.5%
  triton_mm_2875 0.0286 ms 78.3%
SingleProcess AUTOTUNE benchmarking takes 7.9051 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  bias_addmm 1.8341 ms 100.0%
  triton_mm_2898 1.8673 ms 98.2%
  triton_mm_2897 1.9058 ms 96.2%
  triton_mm_2899 2.0540 ms 89.3%
  addmm 2.1823 ms 84.0%
  triton_mm_2892 2.2900 ms 80.1%
  triton_mm_2894 2.3577 ms 77.8%
  triton_mm_2890 2.4138 ms 76.0%
  triton_mm_2893 2.4834 ms 73.9%
  triton_mm_2895 2.6682 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 12.0123 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  mm 1.6968 ms 100.0%
  triton_mm_2935 1.7592 ms 96.5%
  triton_mm_2936 1.7898 ms 94.8%
  triton_mm_2937 1.9240 ms 88.2%
  triton_mm_2928 2.2418 ms 75.7%
  triton_mm_2930 2.2832 ms 74.3%
  triton_mm_2932 2.3514 ms 72.2%
  triton_mm_2931 2.3643 ms 71.8%
  triton_mm_2933 2.4413 ms 69.5%
  triton_mm_2929 2.4763 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 9.9622 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.7145 ms 100.0%
  triton_mm_2916 1.8660 ms 91.9%
  triton_mm_2917 1.8800 ms 91.2%
  triton_mm_2918 1.8915 ms 90.6%
  triton_mm_2912 2.3402 ms 73.3%
  triton_mm_2911 2.4169 ms 70.9%
  triton_mm_2909 2.4748 ms 69.3%
  triton_mm_2913 2.4997 ms 68.6%
  triton_mm_2910 2.5550 ms 67.1%
  triton_mm_2914 2.6379 ms 65.0%
SingleProcess AUTOTUNE benchmarking takes 9.8068 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4217 ms 100.0%
  mm 0.4255 ms 99.1%
  triton_mm_3007 0.4336 ms 97.3%
  triton_mm_3008 0.4583 ms 92.0%
  triton_mm_3003 0.5222 ms 80.8%
  triton_mm_2999 0.5319 ms 79.3%
  triton_mm_3001 0.5373 ms 78.5%
  triton_mm_3004 0.5625 ms 75.0%
  triton_mm_3002 0.5674 ms 74.3%
  triton_mm_3000 0.5834 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 9.7257 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2961 0.0165 ms 100.0%
  triton_mm_2963 0.0166 ms 99.4%
  triton_mm_2960 0.0172 ms 95.9%
  triton_mm_2962 0.0172 ms 95.7%
  triton_mm_2966 0.0174 ms 94.8%
  triton_mm_2964 0.0175 ms 94.5%
  triton_mm_2967 0.0175 ms 94.3%
  triton_mm_2965 0.0177 ms 93.5%
  triton_mm_2959 0.0181 ms 91.2%
  triton_mm_2969 0.0188 ms 88.1%
SingleProcess AUTOTUNE benchmarking takes 5.4129 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0228 ms 100.0%
  triton_mm_2942 0.0240 ms 94.8%
  triton_mm_2950 0.0246 ms 92.6%
  mm 0.0279 ms 81.6%
  triton_mm_2954 0.0295 ms 77.1%
  triton_mm_2940 0.0351 ms 64.8%
  triton_mm_2941 0.0354 ms 64.3%
  triton_mm_2945 0.0366 ms 62.2%
  triton_mm_2949 0.0397 ms 57.3%
  triton_mm_2939 0.0415 ms 54.8%
SingleProcess AUTOTUNE benchmarking takes 6.5824 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4152 ms 100.0%
  triton_mm_2988 0.4191 ms 99.1%
  mm 0.4393 ms 94.5%
  triton_mm_2989 0.4552 ms 91.2%
  triton_mm_2982 0.5169 ms 80.3%
  triton_mm_2984 0.5263 ms 78.9%
  triton_mm_2980 0.5270 ms 78.8%
  triton_mm_2983 0.5707 ms 72.8%
  triton_mm_2981 0.5880 ms 70.6%
  triton_mm_2985 0.6005 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 10.1991 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4163 ms 100.0%
  triton_mm_3044 0.4189 ms 99.4%
  triton_mm_3045 0.4364 ms 95.4%
  triton_mm_3046 0.4623 ms 90.0%
  triton_mm_3037 0.5202 ms 80.0%
  triton_mm_3041 0.5264 ms 79.1%
  triton_mm_3039 0.5273 ms 78.9%
  triton_mm_3040 0.5564 ms 74.8%
  triton_mm_3042 0.5663 ms 73.5%
  triton_mm_3038 0.5726 ms 72.7%
SingleProcess AUTOTUNE benchmarking takes 9.8368 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4003 ms 100.0%
  triton_mm_3025 0.4273 ms 93.7%
  triton_mm_3026 0.4380 ms 91.4%
  triton_mm_3027 0.4438 ms 90.2%
  triton_mm_3020 0.5196 ms 77.1%
  triton_mm_3018 0.5402 ms 74.1%
  triton_mm_3022 0.5407 ms 74.0%
  triton_mm_3021 0.5429 ms 73.7%
  triton_mm_3019 0.5906 ms 67.8%
  triton_mm_3023 0.6137 ms 65.2%
SingleProcess AUTOTUNE benchmarking takes 13.4040 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  triton_mm_3120 0.3236 ms 100.0%
  triton_mm_3121 0.3356 ms 96.4%
  triton_mm_3122 0.3553 ms 91.1%
  mm 0.3680 ms 87.9%
  triton_mm_3113 0.3949 ms 81.9%
  triton_mm_3115 0.3987 ms 81.2%
  triton_mm_3117 0.3988 ms 81.1%
  triton_mm_3116 0.4213 ms 76.8%
  triton_mm_3118 0.4303 ms 75.2%
  triton_mm_3114 0.4341 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 9.7032 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3201 ms 100.0%
  triton_mm_3102 0.3218 ms 99.5%
  mm 0.3301 ms 97.0%
  triton_mm_3103 0.3344 ms 95.7%
  triton_mm_3096 0.3874 ms 82.6%
  triton_mm_3094 0.3993 ms 80.2%
  triton_mm_3097 0.4051 ms 79.0%
  triton_mm_3098 0.4057 ms 78.9%
  triton_mm_3095 0.4426 ms 72.3%
  triton_mm_3099 0.4586 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 9.6223 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1407 ms 100.0%
  triton_mm_3075 0.1462 ms 96.2%
  triton_mm_3077 0.1478 ms 95.2%
  triton_mm_3079 0.1484 ms 94.8%
  triton_mm_3083 0.1489 ms 94.5%
  triton_mm_3084 0.1510 ms 93.2%
  triton_mm_3078 0.1645 ms 85.5%
  triton_mm_3080 0.1672 ms 84.1%
  triton_mm_3076 0.1700 ms 82.8%
  mm 0.1781 ms 79.0%
SingleProcess AUTOTUNE benchmarking takes 9.5999 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1109 ms 100.0%
  triton_mm_3063 0.1126 ms 98.5%
  triton_mm_3064 0.1139 ms 97.4%
  triton_mm_3065 0.1211 ms 91.6%
  triton_mm_3058 0.1349 ms 82.2%
  triton_mm_3060 0.1382 ms 80.3%
  triton_mm_3056 0.1395 ms 79.5%
  triton_mm_3059 0.1483 ms 74.8%
  triton_mm_3057 0.1553 ms 71.4%
  triton_mm_3061 0.1608 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.4488 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3152 0.0319 ms 100.0%
  triton_mm_3145 0.0322 ms 99.1%
  triton_mm_3147 0.0323 ms 98.9%
  triton_mm_3151 0.0326 ms 97.8%
  triton_mm_3150 0.0332 ms 96.2%
  triton_mm_3154 0.0335 ms 95.2%
  triton_mm_3149 0.0336 ms 95.1%
  triton_mm_3148 0.0345 ms 92.7%
  triton_mm_3146 0.0348 ms 91.7%
  triton_mm_3155 0.0351 ms 91.0%
SingleProcess AUTOTUNE benchmarking takes 5.2513 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4259 ms 100.0%
  mm 0.4337 ms 98.2%
  triton_mm_3192 0.4398 ms 96.8%
  triton_mm_3193 0.4653 ms 91.5%
  triton_mm_3188 0.5328 ms 79.9%
  triton_mm_3184 0.5382 ms 79.1%
  triton_mm_3186 0.5459 ms 78.0%
  triton_mm_3187 0.5787 ms 73.6%
  triton_mm_3189 0.5805 ms 73.4%
  triton_mm_3185 0.5951 ms 71.6%
SingleProcess AUTOTUNE benchmarking takes 2.5669 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0530 ms 100.0%
  triton_mm_3131 0.0530 ms 99.9%
  triton_mm_3127 0.0551 ms 96.2%
  triton_mm_3135 0.0588 ms 90.2%
  triton_mm_3139 0.0692 ms 76.6%
  triton_mm_3126 0.0853 ms 62.1%
  triton_mm_3125 0.0889 ms 59.6%
  triton_mm_3130 0.0957 ms 55.3%
  triton_mm_3134 0.1020 ms 51.9%
  triton_mm_3124 0.1044 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 6.5259 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4183 ms 100.0%
  triton_mm_3173 0.4242 ms 98.6%
  mm 0.4424 ms 94.6%
  triton_mm_3174 0.4643 ms 90.1%
  triton_mm_3167 0.5270 ms 79.4%
  triton_mm_3169 0.5308 ms 78.8%
  triton_mm_3165 0.5364 ms 78.0%
  triton_mm_3168 0.5777 ms 72.4%
  triton_mm_3166 0.5980 ms 70.0%
  triton_mm_3170 0.6143 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.5737 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3979 ms 100.0%
  triton_mm_3210 0.4299 ms 92.5%
  triton_mm_3211 0.4431 ms 89.8%
  triton_mm_3212 0.4476 ms 88.9%
  triton_mm_3205 0.5260 ms 75.6%
  triton_mm_3207 0.5458 ms 72.9%
  triton_mm_3203 0.5477 ms 72.6%
  triton_mm_3206 0.5516 ms 72.1%
  triton_mm_3204 0.6011 ms 66.2%
  triton_mm_3208 0.6197 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.5779 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3211 ms 100.0%
  triton_mm_3305 0.3275 ms 98.1%
  triton_mm_3306 0.3399 ms 94.5%
  triton_mm_3307 0.3612 ms 88.9%
  triton_mm_3298 0.4032 ms 79.6%
  triton_mm_3300 0.4056 ms 79.2%
  triton_mm_3302 0.4056 ms 79.2%
  triton_mm_3301 0.4308 ms 74.5%
  triton_mm_3303 0.4367 ms 73.5%
  triton_mm_3299 0.4420 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 2.5296 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3235 ms 100.0%
  triton_mm_3287 0.3257 ms 99.3%
  mm 0.3309 ms 97.7%
  triton_mm_3288 0.3411 ms 94.8%
  triton_mm_3281 0.3945 ms 82.0%
  triton_mm_3279 0.4068 ms 79.5%
  triton_mm_3282 0.4123 ms 78.4%
  triton_mm_3283 0.4130 ms 78.3%
  triton_mm_3280 0.4502 ms 71.8%
  triton_mm_3284 0.4655 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.5437 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1419 ms 100.0%
  triton_mm_3260 0.1467 ms 96.7%
  triton_mm_3262 0.1483 ms 95.7%
  triton_mm_3264 0.1491 ms 95.2%
  triton_mm_3268 0.1491 ms 95.2%
  triton_mm_3269 0.1513 ms 93.8%
  triton_mm_3263 0.1658 ms 85.6%
  triton_mm_3265 0.1687 ms 84.1%
  triton_mm_3261 0.1696 ms 83.7%
  mm 0.1794 ms 79.1%
SingleProcess AUTOTUNE benchmarking takes 2.4404 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3337 0.0320 ms 100.0%
  triton_mm_3334 0.0323 ms 99.1%
  triton_mm_3330 0.0323 ms 98.9%
  triton_mm_3335 0.0331 ms 96.5%
  triton_mm_3331 0.0334 ms 95.7%
  triton_mm_3332 0.0337 ms 94.9%
  triton_mm_3336 0.0339 ms 94.2%
  triton_mm_3333 0.0344 ms 92.9%
  triton_mm_3339 0.0348 ms 91.9%
  triton_mm_3340 0.0352 ms 90.9%
SingleProcess AUTOTUNE benchmarking takes 2.0153 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3357 0.4177 ms 100.0%
  triton_mm_3358 0.4252 ms 98.2%
  mm 0.4395 ms 95.1%
  triton_mm_3359 0.4619 ms 90.4%
  triton_mm_3352 0.5269 ms 79.3%
  triton_mm_3354 0.5344 ms 78.2%
  triton_mm_3350 0.5370 ms 77.8%
  triton_mm_3353 0.5782 ms 72.2%
  triton_mm_3351 0.5980 ms 69.9%
  triton_mm_3355 0.6145 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.5686 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3980 ms 100.0%
  triton_mm_3395 0.4325 ms 92.0%
  triton_mm_3396 0.4439 ms 89.7%
  triton_mm_3397 0.4514 ms 88.2%
  triton_mm_3390 0.5261 ms 75.7%
  triton_mm_3392 0.5468 ms 72.8%
  triton_mm_3388 0.5479 ms 72.6%
  triton_mm_3391 0.5513 ms 72.2%
  triton_mm_3389 0.6010 ms 66.2%
  triton_mm_3393 0.6239 ms 63.8%
SingleProcess AUTOTUNE benchmarking takes 2.5830 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3518 0.0332 ms 100.0%
  triton_mm_3522 0.0334 ms 99.2%
  triton_mm_3519 0.0336 ms 98.8%
  triton_mm_3515 0.0337 ms 98.5%
  triton_mm_3517 0.0337 ms 98.5%
  triton_mm_3521 0.0340 ms 97.7%
  triton_mm_3520 0.0345 ms 96.1%
  triton_mm_3524 0.0347 ms 95.6%
  triton_mm_3516 0.0348 ms 95.2%
  triton_mm_3523 0.0351 ms 94.4%
SingleProcess AUTOTUNE benchmarking takes 2.0211 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3542 0.4178 ms 100.0%
  triton_mm_3543 0.4231 ms 98.8%
  mm 0.4360 ms 95.8%
  triton_mm_3544 0.4621 ms 90.4%
  triton_mm_3537 0.5274 ms 79.2%
  triton_mm_3539 0.5344 ms 78.2%
  triton_mm_3535 0.5367 ms 77.8%
  triton_mm_3538 0.5786 ms 72.2%
  triton_mm_3536 0.5983 ms 69.8%
  triton_mm_3540 0.6144 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.5660 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3969 ms 100.0%
  triton_mm_3580 0.4332 ms 91.6%
  triton_mm_3581 0.4437 ms 89.4%
  triton_mm_3582 0.4504 ms 88.1%
  triton_mm_3575 0.5248 ms 75.6%
  triton_mm_3577 0.5469 ms 72.6%
  triton_mm_3573 0.5474 ms 72.5%
  triton_mm_3576 0.5521 ms 71.9%
  triton_mm_3574 0.5959 ms 66.6%
  triton_mm_3578 0.6236 ms 63.6%
SingleProcess AUTOTUNE benchmarking takes 2.5860 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3707 0.0320 ms 100.0%
  triton_mm_3704 0.0322 ms 99.3%
  triton_mm_3703 0.0331 ms 96.6%
  triton_mm_3702 0.0337 ms 95.1%
  triton_mm_3700 0.0337 ms 95.0%
  triton_mm_3706 0.0338 ms 94.6%
  triton_mm_3705 0.0344 ms 93.0%
  triton_mm_3709 0.0348 ms 91.9%
  triton_mm_3701 0.0348 ms 91.8%
  triton_mm_3710 0.0351 ms 91.1%
SingleProcess AUTOTUNE benchmarking takes 2.0214 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3727 0.4193 ms 100.0%
  triton_mm_3728 0.4278 ms 98.0%
  mm 0.4434 ms 94.6%
  triton_mm_3729 0.4640 ms 90.4%
  triton_mm_3722 0.5284 ms 79.4%
  triton_mm_3724 0.5321 ms 78.8%
  triton_mm_3720 0.5386 ms 77.9%
  triton_mm_3723 0.5797 ms 72.3%
  triton_mm_3721 0.5989 ms 70.0%
  triton_mm_3725 0.6131 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.5722 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0567 ms 100.0%
  mm 0.0583 ms 97.2%
  triton_mm_7428 0.0591 ms 95.9%
  triton_mm_7424 0.0643 ms 88.2%
  triton_mm_7420 0.0655 ms 86.5%
  triton_mm_7422 0.0659 ms 86.0%
  triton_mm_7429 0.0700 ms 81.0%
  triton_mm_7421 0.0708 ms 80.0%
  triton_mm_7425 0.0734 ms 77.2%
  triton_mm_7423 0.0857 ms 66.1%
SingleProcess AUTOTUNE benchmarking takes 10.8271 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1204 ms 100.0%
  triton_mm_7465 0.1228 ms 98.1%
  triton_mm_7466 0.1271 ms 94.7%
  triton_mm_7458 0.1527 ms 78.8%
  triton_mm_7460 0.1539 ms 78.2%
  triton_mm_7462 0.1553 ms 77.5%
  triton_mm_7467 0.1592 ms 75.6%
  triton_mm_7459 0.1644 ms 73.2%
  triton_mm_7461 0.1676 ms 71.8%
  triton_mm_7463 0.1692 ms 71.1%
SingleProcess AUTOTUNE benchmarking takes 10.3482 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1276 ms 100.0%
  triton_mm_7484 0.1463 ms 87.3%
  triton_mm_7485 0.1490 ms 85.7%
  triton_mm_7486 0.1549 ms 82.4%
  triton_mm_7477 0.1615 ms 79.0%
  triton_mm_7481 0.1636 ms 78.0%
  triton_mm_7479 0.1653 ms 77.2%
  triton_mm_7482 0.1683 ms 75.9%
  triton_mm_7480 0.1690 ms 75.5%
  triton_mm_7478 0.1708 ms 74.7%
SingleProcess AUTOTUNE benchmarking takes 12.3854 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1293 ms 100.0%
  triton_mm_7522 0.1435 ms 90.1%
  triton_mm_7519 0.1496 ms 86.4%
  triton_mm_7523 0.1499 ms 86.3%
  triton_mm_7515 0.1529 ms 84.6%
  triton_mm_7524 0.1533 ms 84.4%
  triton_mm_7518 0.1639 ms 78.9%
  triton_mm_7517 0.1654 ms 78.2%
  triton_mm_7520 0.1704 ms 75.9%
  triton_mm_7516 0.1723 ms 75.0%
SingleProcess AUTOTUNE benchmarking takes 9.9610 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0550 ms 100.0%
  triton_mm_7447 0.0610 ms 90.2%
  triton_mm_7441 0.0637 ms 86.4%
  triton_mm_7448 0.0676 ms 81.4%
  triton_mm_7439 0.0697 ms 79.0%
  triton_mm_7443 0.0703 ms 78.3%
  triton_mm_7446 0.0713 ms 77.2%
  triton_mm_7442 0.0718 ms 76.6%
  triton_mm_7440 0.0729 ms 75.5%
  triton_mm_7444 0.0732 ms 75.2%
SingleProcess AUTOTUNE benchmarking takes 9.8577 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1190 ms 100.0%
  triton_mm_7503 0.1210 ms 98.4%
  triton_mm_7504 0.1255 ms 94.8%
  triton_mm_7505 0.1298 ms 91.7%
  triton_mm_7500 0.1465 ms 81.3%
  triton_mm_7496 0.1478 ms 80.5%
  triton_mm_7498 0.1494 ms 79.6%
  triton_mm_7499 0.1585 ms 75.1%
  triton_mm_7497 0.1638 ms 72.6%
  triton_mm_7501 0.1702 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 10.3475 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0302 ms 100.0%
  triton_mm_8637 0.0360 ms 83.8%
  triton_mm_8638 0.0405 ms 74.5%
  triton_mm_8636 0.0408 ms 74.0%
  triton_mm_8642 0.0420 ms 71.8%
  triton_mm_8645 0.0421 ms 71.6%
  triton_mm_8646 0.0422 ms 71.5%
  triton_mm_8639 0.0433 ms 69.7%
  triton_mm_8650 0.0458 ms 65.9%
  triton_mm_8644 0.0463 ms 65.2%
SingleProcess AUTOTUNE benchmarking takes 7.2681 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_11 0.0247 ms 100.0%
  triton_mm_8 0.0249 ms 99.5%
  triton_mm_12 0.0249 ms 99.5%
  triton_mm_7 0.0251 ms 98.5%
  triton_mm_9 0.0256 ms 96.5%
  triton_mm_6 0.0258 ms 95.9%
  triton_mm_4 0.0264 ms 93.8%
  triton_mm_15 0.0268 ms 92.1%
  triton_mm_3 0.0270 ms 91.7%
  triton_mm_2 0.0271 ms 91.2%
SingleProcess AUTOTUNE benchmarking takes 2.3276 seconds and 0.0184 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0474 ms 100.0%
  triton_mm_25 0.0484 ms 97.9%
  triton_mm_33 0.0488 ms 97.2%
  triton_mm_24 0.0492 ms 96.4%
  triton_mm_18 0.0496 ms 95.5%
  triton_mm_28 0.0501 ms 94.6%
  triton_mm_21 0.0513 ms 92.3%
  bias_addmm 0.0516 ms 91.9%
  triton_mm_19 0.0518 ms 91.4%
  triton_mm_20 0.0538 ms 88.0%
SingleProcess AUTOTUNE benchmarking takes 2.2076 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3485 ms 100.0%
  triton_mm_50 0.3570 ms 97.6%
  triton_mm_52 0.3831 ms 91.0%
  bias_addmm 0.4031 ms 86.4%
  triton_mm_45 0.4146 ms 84.1%
  triton_mm_47 0.4283 ms 81.4%
  triton_mm_43 0.4393 ms 79.3%
  triton_mm_46 0.4594 ms 75.9%
  addmm 0.4649 ms 75.0%
  triton_mm_48 0.4916 ms 70.9%
SingleProcess AUTOTUNE benchmarking takes 2.6559 seconds and 0.0168 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1182 ms 100.0%
  triton_mm_69 0.1220 ms 96.9%
  triton_mm_71 0.1289 ms 91.8%
  bias_addmm 0.1316 ms 89.9%
  triton_mm_64 0.1423 ms 83.1%
  triton_mm_66 0.1482 ms 79.8%
  triton_mm_62 0.1488 ms 79.5%
  addmm 0.1496 ms 79.1%
  triton_mm_65 0.1576 ms 75.0%
  triton_mm_67 0.1696 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.5487 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4436 ms 100.0%
  triton_mm_88 0.4504 ms 98.5%
  triton_mm_90 0.4874 ms 91.0%
  bias_addmm 0.5200 ms 85.3%
  triton_mm_83 0.5464 ms 81.2%
  triton_mm_85 0.5628 ms 78.8%
  triton_mm_81 0.5833 ms 76.1%
  addmm 0.5985 ms 74.1%
  triton_mm_84 0.6016 ms 73.7%
  triton_mm_86 0.6464 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6870 seconds and 0.3125 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_108 0.4351 ms 100.0%
  triton_mm_107 0.4557 ms 95.5%
  triton_mm_109 0.4594 ms 94.7%
  addmm 0.4955 ms 87.8%
  bias_addmm 0.4975 ms 87.5%
  triton_mm_102 0.5442 ms 79.9%
  triton_mm_100 0.5665 ms 76.8%
  triton_mm_104 0.5675 ms 76.7%
  triton_mm_103 0.5711 ms 76.2%
  triton_mm_105 0.6357 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.6979 seconds and 0.2086 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_122 0.0466 ms 100.0%
  triton_mm_126 0.0481 ms 96.8%
  triton_mm_118 0.0484 ms 96.2%
  triton_mm_121 0.0492 ms 94.7%
  triton_mm_117 0.0492 ms 94.6%
  triton_mm_111 0.0494 ms 94.2%
  bias_addmm 0.0507 ms 91.8%
  triton_mm_114 0.0513 ms 90.8%
  triton_mm_112 0.0518 ms 89.9%
  triton_mm_113 0.0538 ms 86.6%
SingleProcess AUTOTUNE benchmarking takes 2.2034 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4345 ms 100.0%
  triton_mm_200 0.4535 ms 95.8%
  triton_mm_202 0.4603 ms 94.4%
  addmm 0.4920 ms 88.3%
  bias_addmm 0.4980 ms 87.3%
  triton_mm_195 0.5398 ms 80.5%
  triton_mm_193 0.5617 ms 77.4%
  triton_mm_197 0.5644 ms 77.0%
  triton_mm_196 0.5668 ms 76.7%
  triton_mm_198 0.6317 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7026 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_215 0.0466 ms 100.0%
  triton_mm_211 0.0477 ms 97.7%
  triton_mm_219 0.0481 ms 96.9%
  triton_mm_214 0.0492 ms 94.8%
  triton_mm_204 0.0496 ms 94.0%
  triton_mm_210 0.0502 ms 92.9%
  triton_mm_207 0.0514 ms 90.7%
  bias_addmm 0.0517 ms 90.1%
  triton_mm_205 0.0519 ms 89.8%
  triton_mm_206 0.0539 ms 86.5%
SingleProcess AUTOTUNE benchmarking takes 2.2045 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_294 0.4352 ms 100.0%
  triton_mm_293 0.4548 ms 95.7%
  triton_mm_295 0.4585 ms 94.9%
  addmm 0.4920 ms 88.4%
  bias_addmm 0.4985 ms 87.3%
  triton_mm_288 0.5402 ms 80.6%
  triton_mm_290 0.5636 ms 77.2%
  triton_mm_286 0.5648 ms 77.0%
  triton_mm_289 0.5650 ms 77.0%
  triton_mm_291 0.6301 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.6990 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4289 ms 100.0%
  triton_mm_387 0.4340 ms 98.8%
  triton_mm_386 0.4548 ms 94.3%
  triton_mm_388 0.4565 ms 93.9%
  bias_addmm 0.4945 ms 86.7%
  triton_mm_381 0.5386 ms 79.6%
  triton_mm_379 0.5619 ms 76.3%
  triton_mm_382 0.5634 ms 76.1%
  triton_mm_383 0.5636 ms 76.1%
  triton_mm_384 0.6307 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.6997 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1180 ms 100.0%
  triton_mm_591 0.1209 ms 97.6%
  triton_mm_593 0.1278 ms 92.3%
  triton_mm_586 0.1444 ms 81.7%
  triton_mm_588 0.1489 ms 79.2%
  triton_mm_584 0.1499 ms 78.7%
  bias_addmm 0.1548 ms 76.2%
  addmm 0.1548 ms 76.2%
  triton_mm_587 0.1636 ms 72.1%
  triton_mm_589 0.1676 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 2.5732 seconds and 0.0248 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1233 ms 100.0%
  triton_mm_610 0.1263 ms 97.6%
  bias_addmm 0.1341 ms 91.9%
  triton_mm_612 0.1343 ms 91.8%
  triton_mm_607 0.1532 ms 80.5%
  addmm 0.1539 ms 80.1%
  triton_mm_605 0.1551 ms 79.5%
  triton_mm_603 0.1565 ms 78.8%
  triton_mm_606 0.1728 ms 71.4%
  triton_mm_608 0.1769 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.5746 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0590 ms 100.0%
  triton_mm_624 0.0598 ms 98.7%
  triton_mm_626 0.0609 ms 96.9%
  triton_mm_622 0.0612 ms 96.5%
  triton_mm_630 0.0630 ms 93.7%
  triton_mm_629 0.0630 ms 93.7%
  triton_mm_631 0.0670 ms 88.1%
  triton_mm_625 0.0675 ms 87.4%
  triton_mm_627 0.0711 ms 83.0%
  triton_mm_623 0.0733 ms 80.5%
SingleProcess AUTOTUNE benchmarking takes 2.3776 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0213 ms 100.0%
  triton_mm_2880 0.0227 ms 93.9%
  triton_mm_2872 0.0227 ms 93.7%
  triton_mm_2868 0.0233 ms 91.3%
  bias_addmm 0.0250 ms 85.0%
  triton_mm_2871 0.0264 ms 80.8%
  triton_mm_2867 0.0274 ms 77.7%
  triton_mm_2866 0.0276 ms 77.1%
  triton_mm_2865 0.0277 ms 76.8%
  triton_mm_2875 0.0283 ms 75.4%
SingleProcess AUTOTUNE benchmarking takes 8.0419 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2898 1.8524 ms 100.0%
  triton_mm_2897 1.8684 ms 99.1%
  bias_addmm 1.9815 ms 93.5%
  triton_mm_2899 2.0543 ms 90.2%
  addmm 2.1734 ms 85.2%
  triton_mm_2892 2.2657 ms 81.8%
  triton_mm_2894 2.3268 ms 79.6%
  triton_mm_2890 2.3887 ms 77.6%
  triton_mm_2893 2.4589 ms 75.3%
  triton_mm_2895 2.6816 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 11.9254 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  mm 1.7185 ms 100.0%
  triton_mm_2935 1.7522 ms 98.1%
  triton_mm_2936 1.8131 ms 94.8%
  triton_mm_2937 1.9284 ms 89.1%
  triton_mm_2928 2.2570 ms 76.1%
  triton_mm_2930 2.3379 ms 73.5%
  triton_mm_2932 2.3672 ms 72.6%
  triton_mm_2931 2.3992 ms 71.6%
  triton_mm_2933 2.4756 ms 69.4%
  triton_mm_2929 2.4766 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 9.9363 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.7219 ms 100.0%
  triton_mm_2916 1.8804 ms 91.6%
  triton_mm_2917 1.9074 ms 90.3%
  triton_mm_2918 1.9141 ms 90.0%
  triton_mm_2912 2.3651 ms 72.8%
  triton_mm_2911 2.4618 ms 69.9%
  triton_mm_2909 2.4990 ms 68.9%
  triton_mm_2910 2.5730 ms 66.9%
  triton_mm_2913 2.5833 ms 66.7%
  triton_mm_2914 2.6715 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 12.8290 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4266 ms 100.0%
  triton_mm_3007 0.4401 ms 96.9%
  triton_mm_3008 0.4645 ms 91.8%
  mm 0.4775 ms 89.3%
  triton_mm_3003 0.5296 ms 80.5%
  triton_mm_2999 0.5416 ms 78.8%
  triton_mm_3001 0.5454 ms 78.2%
  triton_mm_3004 0.5724 ms 74.5%
  triton_mm_3002 0.5761 ms 74.0%
  triton_mm_3000 0.5873 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 10.0877 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2962 0.0159 ms 100.0%
  triton_mm_2964 0.0164 ms 97.5%
  triton_mm_2960 0.0172 ms 92.7%
  triton_mm_2961 0.0174 ms 91.5%
  triton_mm_2966 0.0175 ms 91.0%
  triton_mm_2967 0.0177 ms 90.2%
  triton_mm_2963 0.0177 ms 89.9%
  triton_mm_2965 0.0177 ms 89.9%
  triton_mm_2959 0.0180 ms 88.3%
  triton_mm_2958 0.0187 ms 85.3%
SingleProcess AUTOTUNE benchmarking takes 2.0127 seconds and 0.0008 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0230 ms 100.0%
  triton_mm_2950 0.0246 ms 93.6%
  triton_mm_2942 0.0252 ms 91.4%
  mm 0.0280 ms 82.3%
  triton_mm_2954 0.0296 ms 77.8%
  triton_mm_2940 0.0351 ms 65.5%
  triton_mm_2941 0.0355 ms 64.8%
  triton_mm_2945 0.0374 ms 61.5%
  triton_mm_2949 0.0389 ms 59.2%
  triton_mm_2939 0.0416 ms 55.4%
SingleProcess AUTOTUNE benchmarking takes 2.0968 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4219 ms 100.0%
  triton_mm_2988 0.4255 ms 99.2%
  mm 0.4465 ms 94.5%
  triton_mm_2989 0.4610 ms 91.5%
  triton_mm_2982 0.5263 ms 80.2%
  triton_mm_2980 0.5358 ms 78.7%
  triton_mm_2984 0.5375 ms 78.5%
  triton_mm_2983 0.5762 ms 73.2%
  triton_mm_2981 0.5957 ms 70.8%
  triton_mm_2985 0.6109 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.3950 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  triton_mm_3044 0.4254 ms 100.0%
  triton_mm_3045 0.4427 ms 96.1%
  triton_mm_3046 0.4704 ms 90.4%
  mm 0.4711 ms 90.3%
  triton_mm_3037 0.5203 ms 81.8%
  triton_mm_3039 0.5304 ms 80.2%
  triton_mm_3041 0.5350 ms 79.5%
  triton_mm_3040 0.5590 ms 76.1%
  triton_mm_3042 0.5744 ms 74.1%
  triton_mm_3038 0.5747 ms 74.0%
SingleProcess AUTOTUNE benchmarking takes 10.0254 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4058 ms 100.0%
  triton_mm_3025 0.4342 ms 93.5%
  triton_mm_3026 0.4452 ms 91.1%
  triton_mm_3027 0.4526 ms 89.7%
  triton_mm_3020 0.5276 ms 76.9%
  triton_mm_3018 0.5413 ms 75.0%
  triton_mm_3022 0.5482 ms 74.0%
  triton_mm_3021 0.5526 ms 73.4%
  triton_mm_3019 0.5999 ms 67.6%
  triton_mm_3023 0.6226 ms 65.2%
SingleProcess AUTOTUNE benchmarking takes 2.5776 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3217 ms 100.0%
  triton_mm_3120 0.3251 ms 99.0%
  triton_mm_3121 0.3375 ms 95.3%
  triton_mm_3122 0.3571 ms 90.1%
  triton_mm_3113 0.3984 ms 80.7%
  triton_mm_3115 0.4016 ms 80.1%
  triton_mm_3116 0.4244 ms 75.8%
  triton_mm_3118 0.4336 ms 74.2%
  triton_mm_3114 0.4385 ms 73.4%
  triton_mm_3117 0.4639 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 12.7215 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3222 ms 100.0%
  triton_mm_3102 0.3239 ms 99.5%
  mm 0.3305 ms 97.5%
  triton_mm_3103 0.3380 ms 95.3%
  triton_mm_3096 0.3907 ms 82.5%
  triton_mm_3094 0.4027 ms 80.0%
  triton_mm_3098 0.4072 ms 79.1%
  triton_mm_3097 0.4093 ms 78.7%
  triton_mm_3095 0.4464 ms 72.2%
  triton_mm_3099 0.4634 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 9.7751 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1401 ms 100.0%
  triton_mm_3075 0.1464 ms 95.7%
  triton_mm_3079 0.1478 ms 94.8%
  triton_mm_3077 0.1479 ms 94.7%
  triton_mm_3083 0.1483 ms 94.5%
  triton_mm_3084 0.1494 ms 93.7%
  triton_mm_3078 0.1643 ms 85.3%
  triton_mm_3080 0.1673 ms 83.7%
  triton_mm_3076 0.1701 ms 82.4%
  mm 0.1774 ms 79.0%
SingleProcess AUTOTUNE benchmarking takes 9.7722 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1102 ms 100.0%
  triton_mm_3063 0.1123 ms 98.1%
  triton_mm_3064 0.1147 ms 96.1%
  triton_mm_3065 0.1206 ms 91.4%
  triton_mm_3058 0.1342 ms 82.1%
  triton_mm_3056 0.1392 ms 79.2%
  triton_mm_3060 0.1395 ms 79.0%
  triton_mm_3059 0.1480 ms 74.5%
  triton_mm_3057 0.1556 ms 70.8%
  triton_mm_3061 0.1612 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 10.0949 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3147 0.0323 ms 100.0%
  triton_mm_3151 0.0325 ms 99.5%
  triton_mm_3152 0.0332 ms 97.3%
  triton_mm_3145 0.0333 ms 97.0%
  triton_mm_3149 0.0333 ms 96.9%
  triton_mm_3154 0.0334 ms 96.9%
  triton_mm_3148 0.0343 ms 94.3%
  triton_mm_3150 0.0343 ms 94.3%
  triton_mm_3146 0.0346 ms 93.4%
  triton_mm_3155 0.0351 ms 92.0%
SingleProcess AUTOTUNE benchmarking takes 5.3190 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4247 ms 100.0%
  mm 0.4316 ms 98.4%
  triton_mm_3192 0.4394 ms 96.7%
  triton_mm_3193 0.4648 ms 91.4%
  triton_mm_3188 0.5292 ms 80.3%
  triton_mm_3184 0.5399 ms 78.7%
  triton_mm_3186 0.5449 ms 77.9%
  triton_mm_3189 0.5727 ms 74.2%
  triton_mm_3187 0.5786 ms 73.4%
  triton_mm_3185 0.5881 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 2.5691 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0532 ms 100.0%
  triton_mm_3127 0.0532 ms 99.9%
  triton_mm_3131 0.0533 ms 99.6%
  triton_mm_3135 0.0577 ms 92.1%
  triton_mm_3139 0.0692 ms 76.8%
  triton_mm_3126 0.0849 ms 62.6%
  triton_mm_3125 0.0887 ms 59.9%
  triton_mm_3130 0.0949 ms 56.0%
  triton_mm_3134 0.1012 ms 52.5%
  triton_mm_3124 0.1040 ms 51.1%
SingleProcess AUTOTUNE benchmarking takes 6.6684 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4176 ms 100.0%
  triton_mm_3173 0.4249 ms 98.3%
  mm 0.4448 ms 93.9%
  triton_mm_3174 0.4612 ms 90.5%
  triton_mm_3167 0.5261 ms 79.4%
  triton_mm_3169 0.5352 ms 78.0%
  triton_mm_3165 0.5364 ms 77.8%
  triton_mm_3168 0.5775 ms 72.3%
  triton_mm_3166 0.5972 ms 69.9%
  triton_mm_3170 0.6132 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.5690 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4206 ms 100.0%
  triton_mm_3229 0.4234 ms 99.3%
  triton_mm_3230 0.4429 ms 95.0%
  triton_mm_3231 0.4674 ms 90.0%
  triton_mm_3222 0.5249 ms 80.1%
  triton_mm_3226 0.5339 ms 78.8%
  triton_mm_3224 0.5346 ms 78.7%
  triton_mm_3225 0.5670 ms 74.2%
  triton_mm_3227 0.5749 ms 73.2%
  triton_mm_3223 0.5813 ms 72.4%
SingleProcess AUTOTUNE benchmarking takes 2.5707 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4024 ms 100.0%
  triton_mm_3210 0.4324 ms 93.1%
  triton_mm_3211 0.4455 ms 90.3%
  triton_mm_3212 0.4510 ms 89.2%
  triton_mm_3205 0.5269 ms 76.4%
  triton_mm_3203 0.5377 ms 74.8%
  triton_mm_3207 0.5474 ms 73.5%
  triton_mm_3206 0.5516 ms 73.0%
  triton_mm_3204 0.6013 ms 66.9%
  triton_mm_3208 0.6251 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.5875 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3221 ms 100.0%
  triton_mm_3305 0.3273 ms 98.4%
  triton_mm_3306 0.3420 ms 94.2%
  triton_mm_3307 0.3601 ms 89.4%
  triton_mm_3298 0.4007 ms 80.4%
  triton_mm_3302 0.4040 ms 79.7%
  triton_mm_3300 0.4069 ms 79.2%
  triton_mm_3301 0.4303 ms 74.8%
  triton_mm_3303 0.4371 ms 73.7%
  triton_mm_3299 0.4403 ms 73.2%
SingleProcess AUTOTUNE benchmarking takes 2.5350 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3234 ms 100.0%
  triton_mm_3287 0.3268 ms 99.0%
  mm 0.3294 ms 98.2%
  triton_mm_3288 0.3393 ms 95.3%
  triton_mm_3281 0.3945 ms 82.0%
  triton_mm_3279 0.4062 ms 79.6%
  triton_mm_3283 0.4122 ms 78.5%
  triton_mm_3282 0.4141 ms 78.1%
  triton_mm_3280 0.4498 ms 71.9%
  triton_mm_3284 0.4663 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.5521 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1410 ms 100.0%
  triton_mm_3264 0.1471 ms 95.8%
  triton_mm_3260 0.1473 ms 95.7%
  triton_mm_3268 0.1476 ms 95.5%
  triton_mm_3262 0.1480 ms 95.2%
  triton_mm_3269 0.1496 ms 94.2%
  triton_mm_3263 0.1651 ms 85.4%
  triton_mm_3265 0.1684 ms 83.7%
  triton_mm_3261 0.1696 ms 83.1%
  mm 0.1781 ms 79.1%
SingleProcess AUTOTUNE benchmarking takes 2.4383 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1112 ms 100.0%
  triton_mm_3248 0.1117 ms 99.6%
  triton_mm_3249 0.1147 ms 97.0%
  triton_mm_3250 0.1217 ms 91.4%
  triton_mm_3243 0.1346 ms 82.6%
  triton_mm_3241 0.1392 ms 79.9%
  triton_mm_3245 0.1413 ms 78.7%
  triton_mm_3244 0.1476 ms 75.3%
  triton_mm_3242 0.1556 ms 71.4%
  triton_mm_3246 0.1615 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.4282 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3337 0.0321 ms 100.0%
  triton_mm_3335 0.0333 ms 96.4%
  triton_mm_3339 0.0334 ms 96.1%
  triton_mm_3332 0.0336 ms 95.7%
  triton_mm_3330 0.0337 ms 95.4%
  triton_mm_3334 0.0337 ms 95.4%
  triton_mm_3336 0.0338 ms 95.2%
  triton_mm_3333 0.0343 ms 93.6%
  triton_mm_3331 0.0348 ms 92.4%
  triton_mm_3340 0.0352 ms 91.2%
SingleProcess AUTOTUNE benchmarking takes 2.0247 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3376 0.4268 ms 100.0%
  mm 0.4293 ms 99.4%
  triton_mm_3377 0.4403 ms 96.9%
  triton_mm_3378 0.4648 ms 91.8%
  triton_mm_3373 0.5299 ms 80.5%
  triton_mm_3369 0.5436 ms 78.5%
  triton_mm_3371 0.5449 ms 78.3%
  triton_mm_3374 0.5740 ms 74.4%
  triton_mm_3372 0.5809 ms 73.5%
  triton_mm_3370 0.5900 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 2.5734 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3316 0.0525 ms 100.0%
  mm 0.0531 ms 98.9%
  triton_mm_3312 0.0534 ms 98.4%
  triton_mm_3320 0.0577 ms 91.1%
  triton_mm_3324 0.0681 ms 77.1%
  triton_mm_3311 0.0850 ms 61.8%
  triton_mm_3310 0.0880 ms 59.7%
  triton_mm_3315 0.0956 ms 54.9%
  triton_mm_3319 0.1021 ms 51.5%
  triton_mm_3309 0.1040 ms 50.5%
SingleProcess AUTOTUNE benchmarking takes 2.1401 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4199 ms 100.0%
  triton_mm_3414 0.4236 ms 99.1%
  triton_mm_3415 0.4430 ms 94.8%
  triton_mm_3416 0.4677 ms 89.8%
  triton_mm_3407 0.5245 ms 80.0%
  triton_mm_3409 0.5342 ms 78.6%
  triton_mm_3411 0.5348 ms 78.5%
  triton_mm_3410 0.5658 ms 74.2%
  triton_mm_3412 0.5745 ms 73.1%
  triton_mm_3408 0.5813 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 2.5706 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4021 ms 100.0%
  triton_mm_3395 0.4322 ms 93.0%
  triton_mm_3396 0.4460 ms 90.2%
  triton_mm_3397 0.4526 ms 88.8%
  triton_mm_3390 0.5279 ms 76.2%
  triton_mm_3388 0.5412 ms 74.3%
  triton_mm_3392 0.5473 ms 73.5%
  triton_mm_3391 0.5531 ms 72.7%
  triton_mm_3389 0.6021 ms 66.8%
  triton_mm_3393 0.6263 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.5856 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3225 ms 100.0%
  triton_mm_3490 0.3277 ms 98.4%
  triton_mm_3491 0.3415 ms 94.4%
  triton_mm_3492 0.3621 ms 89.1%
  triton_mm_3483 0.4022 ms 80.2%
  triton_mm_3487 0.4056 ms 79.5%
  triton_mm_3485 0.4069 ms 79.2%
  triton_mm_3486 0.4303 ms 75.0%
  triton_mm_3488 0.4380 ms 73.6%
  triton_mm_3484 0.4412 ms 73.1%
SingleProcess AUTOTUNE benchmarking takes 2.5416 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3471 0.3251 ms 100.0%
  triton_mm_3472 0.3272 ms 99.4%
  mm 0.3308 ms 98.3%
  triton_mm_3473 0.3402 ms 95.6%
  triton_mm_3466 0.3952 ms 82.3%
  triton_mm_3464 0.4060 ms 80.1%
  triton_mm_3468 0.4108 ms 79.1%
  triton_mm_3467 0.4142 ms 78.5%
  triton_mm_3465 0.4511 ms 72.1%
  triton_mm_3469 0.4683 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.5459 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3452 0.1411 ms 100.0%
  triton_mm_3445 0.1473 ms 95.8%
  triton_mm_3453 0.1485 ms 95.0%
  triton_mm_3449 0.1486 ms 95.0%
  triton_mm_3447 0.1494 ms 94.4%
  triton_mm_3454 0.1505 ms 93.8%
  triton_mm_3448 0.1643 ms 85.9%
  triton_mm_3450 0.1685 ms 83.7%
  triton_mm_3446 0.1697 ms 83.2%
  mm 0.1795 ms 78.6%
SingleProcess AUTOTUNE benchmarking takes 2.4432 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1112 ms 100.0%
  triton_mm_3433 0.1118 ms 99.4%
  triton_mm_3434 0.1151 ms 96.5%
  triton_mm_3435 0.1220 ms 91.1%
  triton_mm_3428 0.1356 ms 82.0%
  triton_mm_3426 0.1401 ms 79.3%
  triton_mm_3430 0.1409 ms 78.9%
  triton_mm_3429 0.1484 ms 74.9%
  triton_mm_3427 0.1564 ms 71.1%
  triton_mm_3431 0.1615 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.4426 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0529 ms 100.0%
  triton_mm_3501 0.0534 ms 99.1%
  triton_mm_3497 0.0536 ms 98.8%
  triton_mm_3505 0.0578 ms 91.6%
  triton_mm_3509 0.0692 ms 76.4%
  triton_mm_3496 0.0851 ms 62.2%
  triton_mm_3495 0.0880 ms 60.1%
  triton_mm_3500 0.0957 ms 55.3%
  triton_mm_3504 0.1020 ms 51.9%
  triton_mm_3494 0.1040 ms 50.9%
SingleProcess AUTOTUNE benchmarking takes 2.1397 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3542 0.4197 ms 100.0%
  triton_mm_3543 0.4268 ms 98.3%
  mm 0.4427 ms 94.8%
  triton_mm_3544 0.4631 ms 90.6%
  triton_mm_3537 0.5278 ms 79.5%
  triton_mm_3539 0.5368 ms 78.2%
  triton_mm_3535 0.5388 ms 77.9%
  triton_mm_3538 0.5787 ms 72.5%
  triton_mm_3536 0.5985 ms 70.1%
  triton_mm_3540 0.6101 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.5694 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0529 ms 100.0%
  triton_mm_3686 0.0533 ms 99.3%
  triton_mm_3682 0.0535 ms 99.0%
  triton_mm_3690 0.0576 ms 91.8%
  triton_mm_3694 0.0681 ms 77.7%
  triton_mm_3681 0.0842 ms 62.8%
  triton_mm_3680 0.0889 ms 59.5%
  triton_mm_3685 0.0957 ms 55.3%
  triton_mm_3689 0.1012 ms 52.3%
  triton_mm_3679 0.1033 ms 51.3%
SingleProcess AUTOTUNE benchmarking takes 2.1444 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3727 0.4190 ms 100.0%
  triton_mm_3728 0.4244 ms 98.7%
  mm 0.4464 ms 93.8%
  triton_mm_3729 0.4617 ms 90.7%
  triton_mm_3722 0.5264 ms 79.6%
  triton_mm_3724 0.5357 ms 78.2%
  triton_mm_3720 0.5373 ms 78.0%
  triton_mm_3723 0.5736 ms 73.0%
  triton_mm_3721 0.5936 ms 70.6%
  triton_mm_3725 0.6129 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.5763 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0530 ms 100.0%
  triton_mm_3871 0.0534 ms 99.2%
  triton_mm_3867 0.0536 ms 98.9%
  triton_mm_3875 0.0589 ms 90.0%
  triton_mm_3879 0.0694 ms 76.3%
  triton_mm_3866 0.0851 ms 62.2%
  triton_mm_3865 0.0890 ms 59.5%
  triton_mm_3870 0.0958 ms 55.3%
  triton_mm_3874 0.1022 ms 51.8%
  triton_mm_3864 0.1041 ms 50.9%
SingleProcess AUTOTUNE benchmarking takes 2.1486 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3912 0.4193 ms 100.0%
  triton_mm_3913 0.4243 ms 98.8%
  mm 0.4445 ms 94.3%
  triton_mm_3914 0.4623 ms 90.7%
  triton_mm_3907 0.5278 ms 79.4%
  triton_mm_3909 0.5361 ms 78.2%
  triton_mm_3905 0.5379 ms 78.0%
  triton_mm_3908 0.5780 ms 72.5%
  triton_mm_3906 0.5954 ms 70.4%
  triton_mm_3910 0.6098 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.5779 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_4097 0.4183 ms 100.0%
  triton_mm_4098 0.4258 ms 98.2%
  mm 0.4440 ms 94.2%
  triton_mm_4099 0.4622 ms 90.5%
  triton_mm_4092 0.5235 ms 79.9%
  triton_mm_4090 0.5340 ms 78.3%
  triton_mm_4094 0.5346 ms 78.2%
  triton_mm_4093 0.5751 ms 72.7%
  triton_mm_4091 0.5930 ms 70.5%
  triton_mm_4095 0.6143 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.5808 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_4282 0.4192 ms 100.0%
  triton_mm_4283 0.4250 ms 98.6%
  mm 0.4434 ms 94.5%
  triton_mm_4284 0.4601 ms 91.1%
  triton_mm_4277 0.5245 ms 79.9%
  triton_mm_4275 0.5354 ms 78.3%
  triton_mm_4279 0.5364 ms 78.1%
  triton_mm_4278 0.5735 ms 73.1%
  triton_mm_4276 0.5943 ms 70.5%
  triton_mm_4280 0.6072 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.5782 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_4467 0.4192 ms 100.0%
  triton_mm_4468 0.4225 ms 99.2%
  mm 0.4448 ms 94.3%
  triton_mm_4469 0.4603 ms 91.1%
  triton_mm_4462 0.5240 ms 80.0%
  triton_mm_4460 0.5348 ms 78.4%
  triton_mm_4464 0.5360 ms 78.2%
  triton_mm_4463 0.5741 ms 73.0%
  triton_mm_4461 0.5929 ms 70.7%
  triton_mm_4465 0.6124 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.5771 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0555 ms 100.0%
  mm 0.0558 ms 99.3%
  triton_mm_7428 0.0573 ms 96.7%
  triton_mm_7420 0.0636 ms 87.3%
  triton_mm_7424 0.0650 ms 85.3%
  triton_mm_7422 0.0660 ms 84.1%
  triton_mm_7421 0.0689 ms 80.6%
  triton_mm_7429 0.0697 ms 79.6%
  triton_mm_7425 0.0725 ms 76.5%
  triton_mm_7423 0.0838 ms 66.2%
SingleProcess AUTOTUNE benchmarking takes 10.7953 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1174 ms 100.0%
  triton_mm_7465 0.1210 ms 97.0%
  triton_mm_7466 0.1254 ms 93.6%
  triton_mm_7467 0.1318 ms 89.1%
  triton_mm_7458 0.1523 ms 77.1%
  triton_mm_7460 0.1523 ms 77.1%
  triton_mm_7462 0.1559 ms 75.3%
  triton_mm_7459 0.1649 ms 71.2%
  triton_mm_7461 0.1659 ms 70.8%
  triton_mm_7463 0.1682 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 10.3280 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1258 ms 100.0%
  triton_mm_7484 0.1446 ms 87.0%
  triton_mm_7485 0.1477 ms 85.2%
  triton_mm_7486 0.1527 ms 82.4%
  triton_mm_7477 0.1616 ms 77.8%
  triton_mm_7481 0.1618 ms 77.7%
  triton_mm_7479 0.1653 ms 76.1%
  triton_mm_7480 0.1667 ms 75.4%
  triton_mm_7482 0.1690 ms 74.4%
  triton_mm_7478 0.1708 ms 73.6%
SingleProcess AUTOTUNE benchmarking takes 9.9639 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1289 ms 100.0%
  triton_mm_7522 0.1434 ms 89.9%
  triton_mm_7523 0.1482 ms 87.0%
  triton_mm_7519 0.1484 ms 86.9%
  triton_mm_7515 0.1508 ms 85.5%
  triton_mm_7518 0.1623 ms 79.4%
  triton_mm_7517 0.1634 ms 78.9%
  triton_mm_7520 0.1685 ms 76.5%
  triton_mm_7516 0.1714 ms 75.2%
  triton_mm_7524 0.1821 ms 70.8%
SingleProcess AUTOTUNE benchmarking takes 9.8466 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0552 ms 100.0%
  triton_mm_7447 0.0604 ms 91.4%
  triton_mm_7441 0.0642 ms 86.1%
  triton_mm_7448 0.0674 ms 82.0%
  triton_mm_7439 0.0690 ms 80.1%
  triton_mm_7443 0.0700 ms 79.0%
  triton_mm_7446 0.0702 ms 78.7%
  triton_mm_7442 0.0703 ms 78.5%
  triton_mm_7444 0.0723 ms 76.4%
  triton_mm_7440 0.0726 ms 76.1%
SingleProcess AUTOTUNE benchmarking takes 9.8229 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1162 ms 100.0%
  triton_mm_7503 0.1185 ms 98.1%
  triton_mm_7504 0.1245 ms 93.3%
  triton_mm_7505 0.1276 ms 91.1%
  triton_mm_7500 0.1461 ms 79.5%
  triton_mm_7496 0.1470 ms 79.1%
  triton_mm_7498 0.1477 ms 78.7%
  triton_mm_7499 0.1569 ms 74.1%
  triton_mm_7497 0.1640 ms 70.9%
  triton_mm_7501 0.1683 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.3212 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0304 ms 100.0%
  triton_mm_8637 0.0350 ms 87.0%
  triton_mm_8638 0.0397 ms 76.6%
  triton_mm_8636 0.0403 ms 75.5%
  triton_mm_8645 0.0417 ms 73.0%
  triton_mm_8639 0.0424 ms 71.8%
  triton_mm_8642 0.0427 ms 71.3%
  triton_mm_8646 0.0428 ms 71.2%
  triton_mm_8644 0.0455 ms 66.8%
  triton_mm_8650 0.0461 ms 66.0%
SingleProcess AUTOTUNE benchmarking takes 7.1905 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_7 0.0242 ms 100.0%
  triton_mm_9 0.0247 ms 97.8%
  triton_mm_8 0.0248 ms 97.5%
  triton_mm_4 0.0255 ms 94.8%
  triton_mm_11 0.0256 ms 94.3%
  triton_mm_6 0.0257 ms 94.0%
  triton_mm_12 0.0258 ms 93.6%
  triton_mm_3 0.0261 ms 92.6%
  triton_mm_2 0.0262 ms 92.2%
  triton_mm_15 0.0271 ms 89.0%
SingleProcess AUTOTUNE benchmarking takes 2.1407 seconds and 0.0233 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0464 ms 100.0%
  triton_mm_25 0.0478 ms 97.1%
  triton_mm_33 0.0481 ms 96.6%
  triton_mm_28 0.0492 ms 94.3%
  triton_mm_18 0.0501 ms 92.7%
  triton_mm_24 0.0503 ms 92.3%
  triton_mm_21 0.0511 ms 90.9%
  triton_mm_19 0.0515 ms 90.1%
  bias_addmm 0.0518 ms 89.6%
  triton_mm_31 0.0548 ms 84.8%
SingleProcess AUTOTUNE benchmarking takes 2.1935 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3405 ms 100.0%
  triton_mm_50 0.3496 ms 97.4%
  triton_mm_52 0.3754 ms 90.7%
  triton_mm_45 0.4023 ms 84.6%
  bias_addmm 0.4026 ms 84.6%
  addmm 0.4131 ms 82.4%
  triton_mm_47 0.4197 ms 81.1%
  triton_mm_43 0.4304 ms 79.1%
  triton_mm_46 0.4478 ms 76.0%
  triton_mm_48 0.4814 ms 70.7%
SingleProcess AUTOTUNE benchmarking takes 2.6866 seconds and 0.0140 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1166 ms 100.0%
  triton_mm_69 0.1195 ms 97.6%
  triton_mm_71 0.1260 ms 92.6%
  bias_addmm 0.1284 ms 90.8%
  triton_mm_64 0.1388 ms 84.0%
  triton_mm_66 0.1460 ms 79.9%
  triton_mm_62 0.1460 ms 79.8%
  addmm 0.1466 ms 79.5%
  triton_mm_65 0.1554 ms 75.0%
  triton_mm_67 0.1659 ms 70.3%
SingleProcess AUTOTUNE benchmarking takes 2.5519 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4336 ms 100.0%
  triton_mm_88 0.4433 ms 97.8%
  triton_mm_90 0.4785 ms 90.6%
  bias_addmm 0.5186 ms 83.6%
  triton_mm_83 0.5350 ms 81.1%
  triton_mm_85 0.5524 ms 78.5%
  triton_mm_81 0.5707 ms 76.0%
  triton_mm_84 0.5875 ms 73.8%
  triton_mm_86 0.6307 ms 68.7%
  triton_mm_82 0.6412 ms 67.6%
SingleProcess AUTOTUNE benchmarking takes 2.7017 seconds and 0.6243 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_108 0.4296 ms 100.0%
  addmm 0.4334 ms 99.1%
  triton_mm_107 0.4510 ms 95.3%
  triton_mm_109 0.4535 ms 94.7%
  bias_addmm 0.4969 ms 86.5%
  triton_mm_102 0.5364 ms 80.1%
  triton_mm_100 0.5587 ms 76.9%
  triton_mm_104 0.5590 ms 76.9%
  triton_mm_103 0.5614 ms 76.5%
  triton_mm_105 0.6242 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7022 seconds and 0.2159 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_122 0.0464 ms 100.0%
  triton_mm_118 0.0478 ms 97.1%
  triton_mm_126 0.0480 ms 96.5%
  triton_mm_121 0.0492 ms 94.2%
  triton_mm_111 0.0497 ms 93.4%
  triton_mm_117 0.0501 ms 92.5%
  triton_mm_114 0.0512 ms 90.6%
  triton_mm_112 0.0517 ms 89.7%
  bias_addmm 0.0518 ms 89.6%
  triton_mm_124 0.0546 ms 84.9%
SingleProcess AUTOTUNE benchmarking takes 2.2010 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4296 ms 100.0%
  triton_mm_200 0.4507 ms 95.3%
  triton_mm_202 0.4541 ms 94.6%
  addmm 0.4914 ms 87.4%
  bias_addmm 0.4929 ms 87.2%
  triton_mm_195 0.5372 ms 80.0%
  triton_mm_197 0.5585 ms 76.9%
  triton_mm_193 0.5588 ms 76.9%
  triton_mm_196 0.5624 ms 76.4%
  triton_mm_198 0.6247 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7068 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_215 0.0464 ms 100.0%
  triton_mm_211 0.0477 ms 97.3%
  triton_mm_219 0.0486 ms 95.5%
  triton_mm_214 0.0492 ms 94.5%
  triton_mm_204 0.0497 ms 93.5%
  triton_mm_210 0.0501 ms 92.6%
  triton_mm_207 0.0512 ms 90.7%
  bias_addmm 0.0520 ms 89.3%
  triton_mm_205 0.0523 ms 88.7%
  triton_mm_206 0.0543 ms 85.5%
SingleProcess AUTOTUNE benchmarking takes 2.2017 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_294 0.4292 ms 100.0%
  triton_mm_293 0.4495 ms 95.5%
  triton_mm_295 0.4542 ms 94.5%
  addmm 0.4914 ms 87.3%
  bias_addmm 0.4972 ms 86.3%
  triton_mm_288 0.5371 ms 79.9%
  triton_mm_290 0.5590 ms 76.8%
  triton_mm_286 0.5593 ms 76.7%
  triton_mm_289 0.5621 ms 76.4%
  triton_mm_291 0.6246 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7073 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_308 0.0464 ms 100.0%
  triton_mm_304 0.0478 ms 97.2%
  triton_mm_312 0.0480 ms 96.6%
  triton_mm_297 0.0495 ms 93.7%
  triton_mm_303 0.0497 ms 93.4%
  triton_mm_307 0.0500 ms 92.9%
  triton_mm_300 0.0504 ms 92.0%
  bias_addmm 0.0520 ms 89.3%
  triton_mm_298 0.0522 ms 89.0%
  triton_mm_299 0.0545 ms 85.2%
SingleProcess AUTOTUNE benchmarking takes 2.2005 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_401 0.0465 ms 100.0%
  triton_mm_397 0.0478 ms 97.3%
  triton_mm_405 0.0481 ms 96.6%
  triton_mm_400 0.0492 ms 94.3%
  triton_mm_390 0.0497 ms 93.4%
  triton_mm_396 0.0502 ms 92.6%
  triton_mm_393 0.0505 ms 92.1%
  triton_mm_391 0.0515 ms 90.2%
  bias_addmm 0.0520 ms 89.4%
  triton_mm_403 0.0547 ms 85.0%
SingleProcess AUTOTUNE benchmarking takes 2.1914 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_494 0.0464 ms 100.0%
  triton_mm_490 0.0484 ms 95.8%
  triton_mm_498 0.0486 ms 95.4%
  triton_mm_489 0.0492 ms 94.3%
  triton_mm_483 0.0497 ms 93.4%
  triton_mm_493 0.0502 ms 92.5%
  triton_mm_486 0.0504 ms 92.0%
  triton_mm_484 0.0516 ms 89.8%
  bias_addmm 0.0520 ms 89.2%
  triton_mm_485 0.0550 ms 84.3%
SingleProcess AUTOTUNE benchmarking takes 2.1918 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1176 ms 100.0%
  triton_mm_591 0.1211 ms 97.1%
  triton_mm_593 0.1281 ms 91.8%
  triton_mm_586 0.1444 ms 81.4%
  triton_mm_584 0.1495 ms 78.7%
  triton_mm_588 0.1502 ms 78.3%
  bias_addmm 0.1552 ms 75.8%
  triton_mm_587 0.1636 ms 71.9%
  triton_mm_589 0.1687 ms 69.7%
  triton_mm_585 0.1704 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.5054 seconds and 0.0044 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1239 ms 100.0%
  triton_mm_610 0.1252 ms 99.0%
  bias_addmm 0.1341 ms 92.4%
  triton_mm_612 0.1346 ms 92.1%
  addmm 0.1546 ms 80.1%
  triton_mm_603 0.1548 ms 80.1%
  triton_mm_605 0.1558 ms 79.5%
  triton_mm_607 0.1560 ms 79.4%
  triton_mm_606 0.1719 ms 72.1%
  triton_mm_608 0.1760 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 2.5699 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0585 ms 100.0%
  triton_mm_626 0.0601 ms 97.3%
  triton_mm_622 0.0604 ms 96.8%
  triton_mm_624 0.0607 ms 96.3%
  triton_mm_630 0.0628 ms 93.1%
  triton_mm_629 0.0639 ms 91.5%
  triton_mm_631 0.0669 ms 87.4%
  triton_mm_625 0.0684 ms 85.5%
  triton_mm_623 0.0723 ms 80.9%
  triton_mm_627 0.0725 ms 80.7%
SingleProcess AUTOTUNE benchmarking takes 2.3743 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_644 0.0464 ms 100.0%
  triton_mm_640 0.0477 ms 97.3%
  triton_mm_648 0.0481 ms 96.5%
  triton_mm_643 0.0493 ms 94.2%
  triton_mm_633 0.0495 ms 93.7%
  triton_mm_639 0.0495 ms 93.7%
  bias_addmm 0.0510 ms 91.0%
  triton_mm_636 0.0511 ms 90.9%
  triton_mm_634 0.0524 ms 88.7%
  triton_mm_635 0.0544 ms 85.3%
SingleProcess AUTOTUNE benchmarking takes 2.2187 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_737 0.0471 ms 100.0%
  triton_mm_733 0.0478 ms 98.5%
  triton_mm_741 0.0487 ms 96.7%
  triton_mm_726 0.0497 ms 94.9%
  triton_mm_736 0.0500 ms 94.3%
  triton_mm_732 0.0501 ms 94.1%
  triton_mm_729 0.0510 ms 92.4%
  triton_mm_727 0.0515 ms 91.5%
  bias_addmm 0.0518 ms 90.9%
  triton_mm_728 0.0544 ms 86.7%
SingleProcess AUTOTUNE benchmarking takes 2.2194 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_830 0.0464 ms 100.0%
  triton_mm_834 0.0480 ms 96.7%
  triton_mm_826 0.0484 ms 95.8%
  triton_mm_829 0.0492 ms 94.3%
  triton_mm_825 0.0494 ms 94.0%
  triton_mm_819 0.0496 ms 93.6%
  triton_mm_822 0.0512 ms 90.7%
  bias_addmm 0.0522 ms 88.9%
  triton_mm_820 0.0523 ms 88.7%
  triton_mm_821 0.0544 ms 85.3%
SingleProcess AUTOTUNE benchmarking takes 2.2038 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0225 ms 100.0%
  triton_mm_2872 0.0227 ms 98.9%
  triton_mm_2880 0.0228 ms 98.5%
  triton_mm_2868 0.0236 ms 95.4%
  bias_addmm 0.0253 ms 88.7%
  triton_mm_2871 0.0273 ms 82.4%
  triton_mm_2867 0.0276 ms 81.5%
  triton_mm_2875 0.0276 ms 81.2%
  triton_mm_2865 0.0277 ms 81.0%
  triton_mm_2866 0.0287 ms 78.2%
SingleProcess AUTOTUNE benchmarking takes 7.7403 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2898 1.8769 ms 100.0%
  triton_mm_2897 1.9179 ms 97.9%
  bias_addmm 1.9882 ms 94.4%
  triton_mm_2899 2.0652 ms 90.9%
  addmm 2.1896 ms 85.7%
  triton_mm_2892 2.3015 ms 81.6%
  triton_mm_2894 2.3437 ms 80.1%
  triton_mm_2890 2.4200 ms 77.6%
  triton_mm_2893 2.5008 ms 75.1%
  triton_mm_2895 2.6962 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 11.6990 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  triton_mm_2935 1.7687 ms 100.0%
  triton_mm_2936 1.8122 ms 97.6%
  mm 1.8810 ms 94.0%
  triton_mm_2937 1.9211 ms 92.1%
  triton_mm_2928 2.2605 ms 78.2%
  triton_mm_2930 2.3028 ms 76.8%
  triton_mm_2932 2.3523 ms 75.2%
  triton_mm_2931 2.3897 ms 74.0%
  triton_mm_2933 2.4434 ms 72.4%
  triton_mm_2929 2.4717 ms 71.6%
SingleProcess AUTOTUNE benchmarking takes 2.8529 seconds and 0.2111 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.7141 ms 100.0%
  triton_mm_2916 1.8610 ms 92.1%
  triton_mm_2918 1.9002 ms 90.2%
  triton_mm_2917 1.9185 ms 89.3%
  triton_mm_2912 2.3753 ms 72.2%
  triton_mm_2911 2.4659 ms 69.5%
  triton_mm_2909 2.4944 ms 68.7%
  triton_mm_2913 2.5177 ms 68.1%
  triton_mm_2910 2.5627 ms 66.9%
  triton_mm_2914 2.6630 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.8924 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4246 ms 100.0%
  mm 0.4307 ms 98.6%
  triton_mm_3007 0.4381 ms 96.9%
  triton_mm_3008 0.4628 ms 91.7%
  triton_mm_3003 0.5296 ms 80.2%
  triton_mm_2999 0.5392 ms 78.7%
  triton_mm_3001 0.5449 ms 77.9%
  triton_mm_3004 0.5704 ms 74.4%
  triton_mm_3002 0.5778 ms 73.5%
  triton_mm_3000 0.5849 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 9.6901 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2960 0.0160 ms 100.0%
  triton_mm_2962 0.0160 ms 100.0%
  triton_mm_2964 0.0164 ms 97.5%
  triton_mm_2967 0.0164 ms 97.5%
  triton_mm_2961 0.0166 ms 96.5%
  triton_mm_2963 0.0166 ms 96.3%
  triton_mm_2966 0.0174 ms 92.3%
  triton_mm_2965 0.0175 ms 91.4%
  triton_mm_2957 0.0183 ms 87.4%
  triton_mm_2969 0.0187 ms 85.9%
SingleProcess AUTOTUNE benchmarking takes 5.2541 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0217 ms 100.0%
  triton_mm_2942 0.0240 ms 90.3%
  triton_mm_2950 0.0246 ms 87.9%
  mm 0.0279 ms 77.7%
  triton_mm_2954 0.0294 ms 73.6%
  triton_mm_2941 0.0346 ms 62.6%
  triton_mm_2940 0.0351 ms 61.7%
  triton_mm_2945 0.0366 ms 59.2%
  triton_mm_2949 0.0396 ms 54.7%
  triton_mm_2939 0.0414 ms 52.3%
SingleProcess AUTOTUNE benchmarking takes 6.4473 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4180 ms 100.0%
  triton_mm_2988 0.4221 ms 99.0%
  mm 0.4424 ms 94.5%
  triton_mm_2989 0.4599 ms 90.9%
  triton_mm_2982 0.5295 ms 78.9%
  triton_mm_2984 0.5337 ms 78.3%
  triton_mm_2980 0.5369 ms 77.8%
  triton_mm_2983 0.5749 ms 72.7%
  triton_mm_2981 0.5928 ms 70.5%
  triton_mm_2985 0.6074 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 10.1254 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4184 ms 100.0%
  triton_mm_3044 0.4238 ms 98.7%
  triton_mm_3045 0.4402 ms 95.0%
  triton_mm_3046 0.4660 ms 89.8%
  triton_mm_3037 0.5262 ms 79.5%
  triton_mm_3041 0.5328 ms 78.5%
  triton_mm_3039 0.5361 ms 78.0%
  triton_mm_3040 0.5626 ms 74.4%
  triton_mm_3042 0.5744 ms 72.8%
  triton_mm_3038 0.5785 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 10.1025 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4023 ms 100.0%
  triton_mm_3025 0.4312 ms 93.3%
  triton_mm_3026 0.4426 ms 90.9%
  triton_mm_3027 0.4482 ms 89.8%
  triton_mm_3020 0.5266 ms 76.4%
  triton_mm_3018 0.5381 ms 74.8%
  triton_mm_3022 0.5441 ms 73.9%
  triton_mm_3021 0.5515 ms 72.9%
  triton_mm_3019 0.5997 ms 67.1%
  triton_mm_3023 0.6209 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 9.7334 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3242 ms 100.0%
  triton_mm_3120 0.3273 ms 99.1%
  triton_mm_3121 0.3400 ms 95.3%
  triton_mm_3122 0.3585 ms 90.4%
  triton_mm_3113 0.4007 ms 80.9%
  triton_mm_3115 0.4059 ms 79.9%
  triton_mm_3117 0.4061 ms 79.8%
  triton_mm_3116 0.4261 ms 76.1%
  triton_mm_3118 0.4356 ms 74.4%
  triton_mm_3114 0.4399 ms 73.7%
SingleProcess AUTOTUNE benchmarking takes 9.7440 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3231 ms 100.0%
  triton_mm_3102 0.3248 ms 99.5%
  mm 0.3321 ms 97.3%
  triton_mm_3103 0.3389 ms 95.4%
  triton_mm_3096 0.3941 ms 82.0%
  triton_mm_3094 0.4053 ms 79.7%
  triton_mm_3098 0.4101 ms 78.8%
  triton_mm_3097 0.4127 ms 78.3%
  triton_mm_3095 0.4496 ms 71.9%
  triton_mm_3099 0.4658 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 9.6945 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1408 ms 100.0%
  triton_mm_3075 0.1475 ms 95.5%
  triton_mm_3079 0.1482 ms 95.0%
  triton_mm_3083 0.1482 ms 95.0%
  triton_mm_3077 0.1484 ms 94.9%
  triton_mm_3084 0.1506 ms 93.5%
  triton_mm_3078 0.1644 ms 85.6%
  triton_mm_3080 0.1682 ms 83.7%
  triton_mm_3076 0.1703 ms 82.7%
  mm 0.1772 ms 79.4%
SingleProcess AUTOTUNE benchmarking takes 9.6910 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1104 ms 100.0%
  triton_mm_3063 0.1122 ms 98.3%
  triton_mm_3064 0.1139 ms 96.9%
  triton_mm_3065 0.1220 ms 90.5%
  triton_mm_3058 0.1353 ms 81.6%
  triton_mm_3060 0.1388 ms 79.5%
  triton_mm_3056 0.1395 ms 79.1%
  triton_mm_3059 0.1486 ms 74.3%
  triton_mm_3057 0.1560 ms 70.7%
  triton_mm_3061 0.1608 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 13.7213 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3152 0.0320 ms 100.0%
  triton_mm_3145 0.0324 ms 98.8%
  triton_mm_3147 0.0333 ms 96.0%
  triton_mm_3149 0.0333 ms 96.0%
  triton_mm_3151 0.0335 ms 95.5%
  triton_mm_3154 0.0335 ms 95.3%
  triton_mm_3148 0.0342 ms 93.5%
  triton_mm_3150 0.0342 ms 93.5%
  triton_mm_3146 0.0346 ms 92.5%
  triton_mm_3153 0.0352 ms 90.7%
SingleProcess AUTOTUNE benchmarking takes 5.2638 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4221 ms 100.0%
  mm 0.4317 ms 97.8%
  triton_mm_3192 0.4383 ms 96.3%
  triton_mm_3193 0.4628 ms 91.2%
  triton_mm_3188 0.5309 ms 79.5%
  triton_mm_3184 0.5397 ms 78.2%
  triton_mm_3186 0.5441 ms 77.6%
  triton_mm_3189 0.5724 ms 73.7%
  triton_mm_3187 0.5781 ms 73.0%
  triton_mm_3185 0.5835 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 2.5640 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3127 0.0530 ms 100.0%
  triton_mm_3131 0.0530 ms 100.0%
  mm 0.0532 ms 99.8%
  triton_mm_3135 0.0577 ms 92.0%
  triton_mm_3139 0.0690 ms 76.8%
  triton_mm_3126 0.0850 ms 62.4%
  triton_mm_3125 0.0885 ms 59.9%
  triton_mm_3130 0.0954 ms 55.6%
  triton_mm_3134 0.1018 ms 52.1%
  triton_mm_3124 0.1040 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 6.5019 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4183 ms 100.0%
  triton_mm_3229 0.4222 ms 99.1%
  triton_mm_3230 0.4406 ms 94.9%
  triton_mm_3231 0.4663 ms 89.7%
  triton_mm_3222 0.5253 ms 79.6%
  triton_mm_3226 0.5332 ms 78.5%
  triton_mm_3224 0.5337 ms 78.4%
  triton_mm_3225 0.5628 ms 74.3%
  triton_mm_3227 0.5714 ms 73.2%
  triton_mm_3223 0.5770 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.5630 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4013 ms 100.0%
  triton_mm_3210 0.4308 ms 93.2%
  triton_mm_3211 0.4433 ms 90.5%
  triton_mm_3212 0.4515 ms 88.9%
  triton_mm_3205 0.5280 ms 76.0%
  triton_mm_3203 0.5360 ms 74.9%
  triton_mm_3207 0.5459 ms 73.5%
  triton_mm_3206 0.5526 ms 72.6%
  triton_mm_3204 0.5973 ms 67.2%
  triton_mm_3208 0.6247 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.5799 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3246 ms 100.0%
  triton_mm_3305 0.3255 ms 99.7%
  triton_mm_3306 0.3406 ms 95.3%
  triton_mm_3307 0.3595 ms 90.3%
  triton_mm_3298 0.4020 ms 80.8%
  triton_mm_3302 0.4055 ms 80.1%
  triton_mm_3300 0.4066 ms 79.8%
  triton_mm_3301 0.4278 ms 75.9%
  triton_mm_3303 0.4355 ms 74.5%
  triton_mm_3299 0.4412 ms 73.6%
SingleProcess AUTOTUNE benchmarking takes 2.5448 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3225 ms 100.0%
  triton_mm_3287 0.3254 ms 99.1%
  mm 0.3300 ms 97.7%
  triton_mm_3288 0.3390 ms 95.1%
  triton_mm_3281 0.3944 ms 81.8%
  triton_mm_3279 0.4057 ms 79.5%
  triton_mm_3283 0.4095 ms 78.8%
  triton_mm_3282 0.4139 ms 77.9%
  triton_mm_3280 0.4485 ms 71.9%
  triton_mm_3284 0.4645 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.5457 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1398 ms 100.0%
  triton_mm_3268 0.1479 ms 94.6%
  triton_mm_3260 0.1479 ms 94.5%
  triton_mm_3264 0.1484 ms 94.2%
  triton_mm_3262 0.1493 ms 93.6%
  triton_mm_3269 0.1496 ms 93.5%
  triton_mm_3263 0.1642 ms 85.1%
  triton_mm_3265 0.1672 ms 83.6%
  triton_mm_3261 0.1692 ms 82.6%
  mm 0.1795 ms 77.9%
SingleProcess AUTOTUNE benchmarking takes 2.4455 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1099 ms 100.0%
  triton_mm_3248 0.1120 ms 98.1%
  triton_mm_3249 0.1142 ms 96.2%
  triton_mm_3250 0.1208 ms 91.0%
  triton_mm_3243 0.1356 ms 81.0%
  triton_mm_3245 0.1390 ms 79.1%
  triton_mm_3241 0.1392 ms 78.9%
  triton_mm_3244 0.1489 ms 73.8%
  triton_mm_3242 0.1548 ms 71.0%
  triton_mm_3246 0.1603 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.4434 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3337 0.0320 ms 100.0%
  triton_mm_3334 0.0323 ms 99.2%
  triton_mm_3332 0.0324 ms 98.7%
  triton_mm_3330 0.0325 ms 98.5%
  triton_mm_3333 0.0333 ms 96.3%
  triton_mm_3331 0.0335 ms 95.6%
  triton_mm_3336 0.0336 ms 95.3%
  triton_mm_3339 0.0336 ms 95.2%
  triton_mm_3335 0.0342 ms 93.7%
  triton_mm_3340 0.0352 ms 90.9%
SingleProcess AUTOTUNE benchmarking takes 2.0320 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3376 0.4248 ms 100.0%
  mm 0.4331 ms 98.1%
  triton_mm_3377 0.4398 ms 96.6%
  triton_mm_3378 0.4636 ms 91.6%
  triton_mm_3373 0.5343 ms 79.5%
  triton_mm_3369 0.5428 ms 78.3%
  triton_mm_3371 0.5440 ms 78.1%
  triton_mm_3374 0.5692 ms 74.6%
  triton_mm_3372 0.5775 ms 73.6%
  triton_mm_3370 0.5884 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 2.5727 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3316 0.0529 ms 100.0%
  mm 0.0532 ms 99.5%
  triton_mm_3312 0.0536 ms 98.8%
  triton_mm_3320 0.0575 ms 92.0%
  triton_mm_3324 0.0691 ms 76.6%
  triton_mm_3311 0.0850 ms 62.3%
  triton_mm_3310 0.0886 ms 59.7%
  triton_mm_3315 0.0955 ms 55.4%
  triton_mm_3319 0.1018 ms 52.0%
  triton_mm_3309 0.1040 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 2.1518 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0545 ms 100.0%
  mm 0.0551 ms 99.0%
  triton_mm_7428 0.0570 ms 95.6%
  triton_mm_7424 0.0645 ms 84.5%
  triton_mm_7420 0.0646 ms 84.4%
  triton_mm_7422 0.0665 ms 82.0%
  triton_mm_7429 0.0693 ms 78.7%
  triton_mm_7421 0.0697 ms 78.3%
  triton_mm_7425 0.0783 ms 69.6%
  triton_mm_7423 0.0848 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 10.5386 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1181 ms 100.0%
  triton_mm_7466 0.1262 ms 93.5%
  triton_mm_7467 0.1315 ms 89.8%
  triton_mm_7465 0.1417 ms 83.3%
  triton_mm_7462 0.1501 ms 78.7%
  triton_mm_7458 0.1528 ms 77.3%
  triton_mm_7460 0.1528 ms 77.3%
  triton_mm_7459 0.1644 ms 71.8%
  triton_mm_7461 0.1664 ms 70.9%
  triton_mm_7463 0.1677 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 10.2727 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1270 ms 100.0%
  triton_mm_7484 0.1442 ms 88.1%
  triton_mm_7485 0.1492 ms 85.1%
  triton_mm_7486 0.1524 ms 83.3%
  triton_mm_7477 0.1621 ms 78.4%
  triton_mm_7481 0.1622 ms 78.3%
  triton_mm_7479 0.1663 ms 76.4%
  triton_mm_7480 0.1674 ms 75.9%
  triton_mm_7482 0.1692 ms 75.1%
  triton_mm_7478 0.1709 ms 74.3%
SingleProcess AUTOTUNE benchmarking takes 9.7213 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1285 ms 100.0%
  triton_mm_7522 0.1433 ms 89.7%
  triton_mm_7523 0.1477 ms 87.0%
  triton_mm_7519 0.1499 ms 85.7%
  triton_mm_7524 0.1509 ms 85.1%
  triton_mm_7515 0.1525 ms 84.3%
  triton_mm_7518 0.1635 ms 78.6%
  triton_mm_7517 0.1656 ms 77.6%
  triton_mm_7520 0.1686 ms 76.2%
  triton_mm_7516 0.1714 ms 75.0%
SingleProcess AUTOTUNE benchmarking takes 9.8238 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0546 ms 100.0%
  triton_mm_7447 0.0601 ms 90.8%
  triton_mm_7441 0.0639 ms 85.4%
  triton_mm_7448 0.0667 ms 81.8%
  triton_mm_7439 0.0698 ms 78.2%
  triton_mm_7446 0.0700 ms 77.9%
  triton_mm_7442 0.0701 ms 77.9%
  triton_mm_7443 0.0709 ms 76.9%
  triton_mm_7444 0.0720 ms 75.7%
  triton_mm_7440 0.0725 ms 75.2%
SingleProcess AUTOTUNE benchmarking takes 9.7273 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1176 ms 100.0%
  triton_mm_7504 0.1246 ms 94.4%
  triton_mm_7505 0.1276 ms 92.1%
  triton_mm_7503 0.1412 ms 83.3%
  triton_mm_7500 0.1445 ms 81.3%
  triton_mm_7496 0.1482 ms 79.4%
  triton_mm_7498 0.1482 ms 79.3%
  triton_mm_7499 0.1566 ms 75.1%
  triton_mm_7497 0.1644 ms 71.5%
  triton_mm_7501 0.1693 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 10.2519 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0307 ms 100.0%
  triton_mm_8637 0.0354 ms 86.7%
  triton_mm_8636 0.0409 ms 75.1%
  triton_mm_8645 0.0416 ms 73.8%
  triton_mm_8642 0.0427 ms 71.9%
  triton_mm_8646 0.0430 ms 71.4%
  triton_mm_8638 0.0459 ms 66.9%
  triton_mm_8650 0.0461 ms 66.6%
  triton_mm_8644 0.0466 ms 65.8%
  triton_mm_8639 0.0492 ms 62.3%
SingleProcess AUTOTUNE benchmarking takes 9.8380 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_7 0.0241 ms 100.0%
  triton_mm_9 0.0245 ms 98.4%
  triton_mm_11 0.0245 ms 98.4%
  triton_mm_8 0.0246 ms 98.0%
  triton_mm_4 0.0253 ms 95.2%
  triton_mm_6 0.0256 ms 94.3%
  triton_mm_12 0.0258 ms 93.5%
  triton_mm_3 0.0260 ms 92.7%
  triton_mm_2 0.0260 ms 92.6%
  triton_mm_15 0.0278 ms 86.9%
SingleProcess AUTOTUNE benchmarking takes 2.0814 seconds and 0.1070 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0473 ms 100.0%
  triton_mm_33 0.0478 ms 99.0%
  triton_mm_25 0.0482 ms 98.3%
  triton_mm_28 0.0489 ms 96.8%
  triton_mm_24 0.0490 ms 96.5%
  triton_mm_18 0.0494 ms 95.7%
  triton_mm_21 0.0512 ms 92.4%
  bias_addmm 0.0515 ms 91.9%
  triton_mm_19 0.0517 ms 91.5%
  triton_mm_20 0.0540 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.1896 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3464 ms 100.0%
  triton_mm_50 0.3560 ms 97.3%
  triton_mm_52 0.3800 ms 91.2%
  bias_addmm 0.4010 ms 86.4%
  triton_mm_45 0.4103 ms 84.4%
  addmm 0.4205 ms 82.4%
  triton_mm_47 0.4244 ms 81.6%
  triton_mm_43 0.4378 ms 79.1%
  triton_mm_46 0.4589 ms 75.5%
  triton_mm_48 0.4934 ms 70.2%
SingleProcess AUTOTUNE benchmarking takes 2.7027 seconds and 0.0261 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1185 ms 100.0%
  triton_mm_69 0.1216 ms 97.4%
  triton_mm_71 0.1285 ms 92.2%
  bias_addmm 0.1303 ms 90.9%
  triton_mm_64 0.1407 ms 84.2%
  triton_mm_66 0.1467 ms 80.7%
  triton_mm_62 0.1473 ms 80.4%
  addmm 0.1488 ms 79.6%
  triton_mm_65 0.1550 ms 76.4%
  triton_mm_67 0.1677 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.5469 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4412 ms 100.0%
  triton_mm_88 0.4487 ms 98.3%
  triton_mm_90 0.4881 ms 90.4%
  bias_addmm 0.5185 ms 85.1%
  triton_mm_83 0.5496 ms 80.3%
  triton_mm_85 0.5635 ms 78.3%
  triton_mm_81 0.5830 ms 75.7%
  addmm 0.5951 ms 74.1%
  triton_mm_84 0.5990 ms 73.7%
  triton_mm_86 0.6405 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.6908 seconds and 1.6356 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4331 ms 100.0%
  triton_mm_108 0.4348 ms 99.6%
  triton_mm_107 0.4551 ms 95.2%
  triton_mm_109 0.4592 ms 94.3%
  bias_addmm 0.4944 ms 87.6%
  triton_mm_102 0.5437 ms 79.7%
  triton_mm_100 0.5643 ms 76.7%
  triton_mm_104 0.5660 ms 76.5%
  triton_mm_103 0.5687 ms 76.1%
  triton_mm_105 0.6300 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7048 seconds and 0.1865 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_122 0.0466 ms 100.0%
  triton_mm_118 0.0482 ms 96.5%
  triton_mm_126 0.0485 ms 96.0%
  triton_mm_121 0.0489 ms 95.3%
  triton_mm_117 0.0490 ms 95.0%
  triton_mm_111 0.0500 ms 93.1%
  bias_addmm 0.0506 ms 92.0%
  triton_mm_114 0.0512 ms 91.0%
  triton_mm_112 0.0516 ms 90.2%
  triton_mm_113 0.0540 ms 86.2%
SingleProcess AUTOTUNE benchmarking takes 2.1998 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4330 ms 100.0%
  triton_mm_200 0.4536 ms 95.5%
  triton_mm_202 0.4580 ms 94.5%
  addmm 0.4931 ms 87.8%
  bias_addmm 0.4954 ms 87.4%
  triton_mm_195 0.5438 ms 79.6%
  triton_mm_193 0.5637 ms 76.8%
  triton_mm_197 0.5654 ms 76.6%
  triton_mm_196 0.5671 ms 76.4%
  triton_mm_198 0.6292 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.6915 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_215 0.0473 ms 100.0%
  triton_mm_211 0.0482 ms 98.1%
  triton_mm_219 0.0485 ms 97.4%
  triton_mm_210 0.0492 ms 96.2%
  triton_mm_204 0.0494 ms 95.8%
  triton_mm_214 0.0498 ms 95.0%
  triton_mm_207 0.0512 ms 92.5%
  triton_mm_205 0.0517 ms 91.4%
  bias_addmm 0.0519 ms 91.1%
  triton_mm_206 0.0540 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.2142 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_308 0.0473 ms 100.0%
  triton_mm_312 0.0478 ms 99.0%
  triton_mm_304 0.0482 ms 98.1%
  triton_mm_303 0.0489 ms 96.7%
  triton_mm_307 0.0497 ms 95.2%
  triton_mm_297 0.0499 ms 94.8%
  triton_mm_298 0.0510 ms 92.8%
  triton_mm_300 0.0512 ms 92.4%
  bias_addmm 0.0517 ms 91.6%
  triton_mm_299 0.0533 ms 88.7%
SingleProcess AUTOTUNE benchmarking takes 2.2121 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1181 ms 100.0%
  triton_mm_591 0.1216 ms 97.1%
  triton_mm_593 0.1285 ms 91.9%
  triton_mm_586 0.1438 ms 82.1%
  triton_mm_588 0.1478 ms 79.9%
  triton_mm_584 0.1496 ms 78.9%
  bias_addmm 0.1533 ms 77.0%
  triton_mm_587 0.1623 ms 72.7%
  triton_mm_589 0.1668 ms 70.8%
  triton_mm_585 0.1686 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 11.0681 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1238 ms 100.0%
  triton_mm_610 0.1255 ms 98.6%
  triton_mm_612 0.1354 ms 91.5%
  bias_addmm 0.1359 ms 91.1%
  triton_mm_603 0.1520 ms 81.5%
  triton_mm_605 0.1541 ms 80.4%
  triton_mm_607 0.1543 ms 80.2%
  addmm 0.1563 ms 79.2%
  triton_mm_606 0.1729 ms 71.6%
  triton_mm_608 0.1741 ms 71.1%
SingleProcess AUTOTUNE benchmarking takes 2.5788 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0579 ms 100.0%
  triton_mm_626 0.0606 ms 95.6%
  triton_mm_624 0.0606 ms 95.6%
  triton_mm_622 0.0620 ms 93.4%
  triton_mm_630 0.0636 ms 91.1%
  triton_mm_629 0.0639 ms 90.6%
  triton_mm_631 0.0670 ms 86.4%
  triton_mm_625 0.0683 ms 84.8%
  triton_mm_627 0.0724 ms 79.9%
  triton_mm_623 0.0732 ms 79.1%
SingleProcess AUTOTUNE benchmarking takes 2.3767 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0214 ms 100.0%
  triton_mm_2872 0.0227 ms 94.6%
  triton_mm_2868 0.0235 ms 91.4%
  triton_mm_2880 0.0236 ms 90.7%
  bias_addmm 0.0259 ms 82.9%
  triton_mm_2871 0.0272 ms 78.8%
  addmm 0.0274 ms 78.4%
  triton_mm_2867 0.0275 ms 78.1%
  triton_mm_2875 0.0275 ms 78.1%
  triton_mm_2865 0.0276 ms 77.6%
SingleProcess AUTOTUNE benchmarking takes 7.5865 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2897 1.8926 ms 100.0%
  bias_addmm 1.9862 ms 95.3%
  triton_mm_2898 2.0135 ms 94.0%
  triton_mm_2899 2.0632 ms 91.7%
  addmm 2.2125 ms 85.5%
  triton_mm_2892 2.2717 ms 83.3%
  triton_mm_2894 2.3870 ms 79.3%
  triton_mm_2890 2.4448 ms 77.4%
  triton_mm_2895 2.7110 ms 69.8%
  triton_mm_2891 2.7973 ms 67.7%
SingleProcess AUTOTUNE benchmarking takes 12.2715 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  mm 1.7381 ms 100.0%
  triton_mm_2935 1.7625 ms 98.6%
  triton_mm_2936 1.8376 ms 94.6%
  triton_mm_2937 1.9549 ms 88.9%
  triton_mm_2928 2.2910 ms 75.9%
  triton_mm_2930 2.3171 ms 75.0%
  triton_mm_2932 2.3632 ms 73.5%
  triton_mm_2931 2.4340 ms 71.4%
  triton_mm_2933 2.4804 ms 70.1%
  triton_mm_2929 2.5094 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 10.2425 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.7127 ms 100.0%
  triton_mm_2916 1.8578 ms 92.2%
  triton_mm_2917 1.9232 ms 89.1%
  triton_mm_2918 1.9259 ms 88.9%
  triton_mm_2912 2.3865 ms 71.8%
  triton_mm_2911 2.4696 ms 69.3%
  triton_mm_2909 2.4800 ms 69.1%
  triton_mm_2910 2.5956 ms 66.0%
  triton_mm_2913 2.5982 ms 65.9%
  triton_mm_2914 2.6991 ms 63.5%
SingleProcess AUTOTUNE benchmarking takes 10.1890 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4228 ms 100.0%
  triton_mm_3007 0.4361 ms 96.9%
  triton_mm_3008 0.4626 ms 91.4%
  mm 0.4776 ms 88.5%
  triton_mm_3003 0.5284 ms 80.0%
  triton_mm_2999 0.5379 ms 78.6%
  triton_mm_3001 0.5429 ms 77.9%
  triton_mm_3004 0.5700 ms 74.2%
  triton_mm_3002 0.5779 ms 73.2%
  triton_mm_3000 0.5843 ms 72.4%
SingleProcess AUTOTUNE benchmarking takes 9.9834 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2962 0.0160 ms 100.0%
  triton_mm_2963 0.0165 ms 96.5%
  triton_mm_2960 0.0170 ms 94.0%
  triton_mm_2964 0.0170 ms 93.9%
  triton_mm_2966 0.0172 ms 92.9%
  triton_mm_2967 0.0172 ms 92.6%
  triton_mm_2961 0.0174 ms 91.7%
  triton_mm_2965 0.0174 ms 91.6%
  triton_mm_2969 0.0177 ms 90.1%
  triton_mm_2957 0.0181 ms 88.2%
SingleProcess AUTOTUNE benchmarking takes 5.4330 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0225 ms 100.0%
  triton_mm_2942 0.0249 ms 90.5%
  triton_mm_2950 0.0255 ms 88.3%
  mm 0.0279 ms 80.6%
  triton_mm_2954 0.0283 ms 79.4%
  triton_mm_2940 0.0341 ms 65.9%
  triton_mm_2941 0.0351 ms 64.0%
  triton_mm_2945 0.0371 ms 60.7%
  triton_mm_2949 0.0394 ms 57.1%
  triton_mm_2939 0.0412 ms 54.6%
SingleProcess AUTOTUNE benchmarking takes 6.6778 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4207 ms 100.0%
  triton_mm_2988 0.4231 ms 99.4%
  mm 0.4420 ms 95.2%
  triton_mm_2989 0.4648 ms 90.5%
  triton_mm_2982 0.5289 ms 79.5%
  triton_mm_2980 0.5374 ms 78.3%
  triton_mm_2984 0.5379 ms 78.2%
  triton_mm_2983 0.5764 ms 73.0%
  triton_mm_2981 0.6110 ms 68.9%
  triton_mm_2985 0.6130 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.5765 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4178 ms 100.0%
  triton_mm_3044 0.4213 ms 99.2%
  triton_mm_3045 0.4387 ms 95.2%
  triton_mm_3046 0.4642 ms 90.0%
  triton_mm_3037 0.5237 ms 79.8%
  triton_mm_3041 0.5316 ms 78.6%
  triton_mm_3039 0.5329 ms 78.4%
  triton_mm_3040 0.5619 ms 74.4%
  triton_mm_3042 0.5720 ms 73.0%
  triton_mm_3038 0.5781 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 10.0304 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4018 ms 100.0%
  triton_mm_3025 0.4293 ms 93.6%
  triton_mm_3026 0.4416 ms 91.0%
  triton_mm_3027 0.4457 ms 90.2%
  triton_mm_3020 0.5251 ms 76.5%
  triton_mm_3018 0.5362 ms 74.9%
  triton_mm_3022 0.5404 ms 74.4%
  triton_mm_3021 0.5477 ms 73.4%
  triton_mm_3019 0.5979 ms 67.2%
  triton_mm_3023 0.6210 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 9.9894 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3231 ms 100.0%
  triton_mm_3120 0.3255 ms 99.3%
  triton_mm_3121 0.3393 ms 95.2%
  triton_mm_3122 0.3583 ms 90.2%
  triton_mm_3113 0.3988 ms 81.0%
  triton_mm_3117 0.4040 ms 80.0%
  triton_mm_3115 0.4047 ms 79.9%
  triton_mm_3116 0.4257 ms 75.9%
  triton_mm_3118 0.4349 ms 74.3%
  triton_mm_3114 0.4394 ms 73.5%
SingleProcess AUTOTUNE benchmarking takes 11.7993 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3215 ms 100.0%
  triton_mm_3102 0.3239 ms 99.3%
  mm 0.3327 ms 96.7%
  triton_mm_3103 0.3369 ms 95.4%
  triton_mm_3096 0.3925 ms 81.9%
  triton_mm_3094 0.4030 ms 79.8%
  triton_mm_3098 0.4084 ms 78.7%
  triton_mm_3097 0.4093 ms 78.6%
  triton_mm_3095 0.4479 ms 71.8%
  triton_mm_3099 0.4632 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 10.1087 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1415 ms 100.0%
  triton_mm_3075 0.1469 ms 96.3%
  triton_mm_3079 0.1475 ms 95.9%
  triton_mm_3077 0.1481 ms 95.5%
  triton_mm_3083 0.1486 ms 95.2%
  triton_mm_3084 0.1509 ms 93.8%
  triton_mm_3078 0.1645 ms 86.0%
  triton_mm_3080 0.1685 ms 84.0%
  triton_mm_3076 0.1703 ms 83.1%
  mm 0.1786 ms 79.2%
SingleProcess AUTOTUNE benchmarking takes 10.1568 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1115 ms 100.0%
  triton_mm_3063 0.1122 ms 99.4%
  triton_mm_3064 0.1142 ms 97.6%
  triton_mm_3065 0.1213 ms 91.9%
  triton_mm_3058 0.1346 ms 82.8%
  triton_mm_3060 0.1399 ms 79.7%
  triton_mm_3056 0.1400 ms 79.6%
  triton_mm_3059 0.1490 ms 74.8%
  triton_mm_3057 0.1548 ms 72.0%
  triton_mm_3061 0.1612 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 10.5245 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3152 0.0319 ms 100.0%
  triton_mm_3149 0.0321 ms 99.6%
  triton_mm_3145 0.0332 ms 96.1%
  triton_mm_3147 0.0333 ms 96.0%
  triton_mm_3151 0.0335 ms 95.4%
  triton_mm_3148 0.0340 ms 94.0%
  triton_mm_3150 0.0340 ms 94.0%
  triton_mm_3146 0.0345 ms 92.6%
  triton_mm_3154 0.0345 ms 92.5%
  triton_mm_3153 0.0361 ms 88.5%
SingleProcess AUTOTUNE benchmarking takes 5.4118 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4244 ms 100.0%
  mm 0.4302 ms 98.7%
  triton_mm_3192 0.4386 ms 96.8%
  triton_mm_3193 0.4636 ms 91.6%
  triton_mm_3188 0.5295 ms 80.2%
  triton_mm_3184 0.5423 ms 78.3%
  triton_mm_3186 0.5495 ms 77.2%
  triton_mm_3189 0.5748 ms 73.8%
  triton_mm_3187 0.5807 ms 73.1%
  triton_mm_3185 0.5904 ms 71.9%
SingleProcess AUTOTUNE benchmarking takes 2.5720 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3127 0.0530 ms 100.0%
  mm 0.0530 ms 99.9%
  triton_mm_3131 0.0531 ms 99.8%
  triton_mm_3135 0.0574 ms 92.3%
  triton_mm_3139 0.0691 ms 76.7%
  triton_mm_3126 0.0848 ms 62.5%
  triton_mm_3125 0.0883 ms 60.0%
  triton_mm_3130 0.0951 ms 55.7%
  triton_mm_3134 0.1009 ms 52.5%
  triton_mm_3124 0.1031 ms 51.4%
SingleProcess AUTOTUNE benchmarking takes 6.6743 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4214 ms 100.0%
  triton_mm_3173 0.4272 ms 98.7%
  mm 0.4374 ms 96.4%
  triton_mm_3174 0.4668 ms 90.3%
  triton_mm_3167 0.5310 ms 79.4%
  triton_mm_3169 0.5395 ms 78.1%
  triton_mm_3165 0.5404 ms 78.0%
  triton_mm_3168 0.5812 ms 72.5%
  triton_mm_3166 0.6012 ms 70.1%
  triton_mm_3170 0.6160 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.5746 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3235 ms 100.0%
  triton_mm_3305 0.3274 ms 98.8%
  triton_mm_3306 0.3396 ms 95.2%
  triton_mm_3307 0.3595 ms 90.0%
  triton_mm_3298 0.4009 ms 80.7%
  triton_mm_3302 0.4042 ms 80.0%
  triton_mm_3300 0.4050 ms 79.9%
  triton_mm_3301 0.4289 ms 75.4%
  triton_mm_3303 0.4370 ms 74.0%
  triton_mm_3299 0.4416 ms 73.3%
SingleProcess AUTOTUNE benchmarking takes 2.5505 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3228 ms 100.0%
  triton_mm_3287 0.3248 ms 99.4%
  mm 0.3303 ms 97.7%
  triton_mm_3288 0.3398 ms 95.0%
  triton_mm_3281 0.3931 ms 82.1%
  triton_mm_3279 0.4049 ms 79.7%
  triton_mm_3283 0.4101 ms 78.7%
  triton_mm_3282 0.4153 ms 77.7%
  triton_mm_3280 0.4511 ms 71.6%
  triton_mm_3284 0.4684 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.5588 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1408 ms 100.0%
  triton_mm_3264 0.1476 ms 95.4%
  triton_mm_3260 0.1477 ms 95.3%
  triton_mm_3268 0.1486 ms 94.8%
  triton_mm_3262 0.1488 ms 94.6%
  triton_mm_3269 0.1510 ms 93.3%
  triton_mm_3263 0.1650 ms 85.3%
  triton_mm_3265 0.1676 ms 84.0%
  triton_mm_3261 0.1700 ms 82.8%
  mm 0.1796 ms 78.4%
SingleProcess AUTOTUNE benchmarking takes 2.4542 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1115 ms 100.0%
  triton_mm_3248 0.1121 ms 99.4%
  triton_mm_3249 0.1148 ms 97.1%
  triton_mm_3250 0.1218 ms 91.5%
  triton_mm_3243 0.1356 ms 82.2%
  triton_mm_3241 0.1390 ms 80.2%
  triton_mm_3245 0.1397 ms 79.8%
  triton_mm_3244 0.1487 ms 75.0%
  triton_mm_3242 0.1544 ms 72.2%
  triton_mm_3246 0.1608 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.4514 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3337 0.0320 ms 100.0%
  triton_mm_3336 0.0325 ms 98.3%
  triton_mm_3330 0.0332 ms 96.3%
  triton_mm_3334 0.0332 ms 96.3%
  triton_mm_3335 0.0332 ms 96.3%
  triton_mm_3332 0.0333 ms 96.2%
  triton_mm_3333 0.0333 ms 96.1%
  triton_mm_3339 0.0335 ms 95.6%
  triton_mm_3331 0.0345 ms 92.8%
  triton_mm_3340 0.0351 ms 91.2%
SingleProcess AUTOTUNE benchmarking takes 2.0240 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0528 ms 100.0%
  triton_mm_3316 0.0529 ms 99.7%
  triton_mm_3312 0.0537 ms 98.2%
  triton_mm_3320 0.0583 ms 90.5%
  triton_mm_3324 0.0681 ms 77.5%
  triton_mm_3311 0.0842 ms 62.7%
  triton_mm_3310 0.0877 ms 60.1%
  triton_mm_3315 0.0952 ms 55.4%
  triton_mm_3319 0.1010 ms 52.2%
  triton_mm_3309 0.1031 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 2.1490 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3357 0.4208 ms 100.0%
  triton_mm_3358 0.4259 ms 98.8%
  mm 0.4387 ms 95.9%
  triton_mm_3359 0.4640 ms 90.7%
  triton_mm_3352 0.5355 ms 78.6%
  triton_mm_3354 0.5386 ms 78.1%
  triton_mm_3350 0.5390 ms 78.1%
  triton_mm_3353 0.5798 ms 72.6%
  triton_mm_3351 0.5980 ms 70.4%
  triton_mm_3355 0.6133 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.5722 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3205 ms 100.0%
  triton_mm_3490 0.3267 ms 98.1%
  triton_mm_3491 0.3407 ms 94.1%
  triton_mm_3492 0.3610 ms 88.8%
  triton_mm_3483 0.4009 ms 80.0%
  triton_mm_3487 0.4053 ms 79.1%
  triton_mm_3485 0.4057 ms 79.0%
  triton_mm_3486 0.4322 ms 74.2%
  triton_mm_3488 0.4372 ms 73.3%
  triton_mm_3484 0.4418 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 2.5497 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3471 0.3232 ms 100.0%
  triton_mm_3472 0.3260 ms 99.1%
  mm 0.3314 ms 97.5%
  triton_mm_3473 0.3413 ms 94.7%
  triton_mm_3466 0.3937 ms 82.1%
  triton_mm_3464 0.4061 ms 79.6%
  triton_mm_3467 0.4108 ms 78.7%
  triton_mm_3468 0.4111 ms 78.6%
  triton_mm_3465 0.4508 ms 71.7%
  triton_mm_3469 0.4676 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.5559 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3452 0.1409 ms 100.0%
  triton_mm_3445 0.1486 ms 94.9%
  triton_mm_3449 0.1486 ms 94.8%
  triton_mm_3453 0.1487 ms 94.8%
  triton_mm_3447 0.1495 ms 94.3%
  triton_mm_3454 0.1512 ms 93.2%
  triton_mm_3448 0.1649 ms 85.5%
  triton_mm_3450 0.1678 ms 84.0%
  triton_mm_3446 0.1704 ms 82.7%
  mm 0.1807 ms 78.0%
SingleProcess AUTOTUNE benchmarking takes 2.4545 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1107 ms 100.0%
  triton_mm_3433 0.1127 ms 98.2%
  triton_mm_3434 0.1152 ms 96.1%
  triton_mm_3435 0.1224 ms 90.4%
  triton_mm_3428 0.1363 ms 81.2%
  triton_mm_3430 0.1403 ms 78.9%
  triton_mm_3426 0.1404 ms 78.8%
  triton_mm_3429 0.1492 ms 74.1%
  triton_mm_3427 0.1560 ms 70.9%
  triton_mm_3431 0.1616 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.4488 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3501 0.0528 ms 100.0%
  mm 0.0529 ms 99.8%
  triton_mm_3497 0.0534 ms 98.9%
  triton_mm_3505 0.0574 ms 92.1%
  triton_mm_3509 0.0691 ms 76.4%
  triton_mm_3496 0.0842 ms 62.7%
  triton_mm_3495 0.0877 ms 60.2%
  triton_mm_3500 0.0951 ms 55.6%
  triton_mm_3504 0.1016 ms 52.0%
  triton_mm_3494 0.1031 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 2.1520 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3542 0.4197 ms 100.0%
  triton_mm_3543 0.4258 ms 98.6%
  mm 0.4382 ms 95.8%
  triton_mm_3544 0.4639 ms 90.5%
  triton_mm_3537 0.5314 ms 79.0%
  triton_mm_3539 0.5410 ms 77.6%
  triton_mm_3535 0.5413 ms 77.5%
  triton_mm_3538 0.5813 ms 72.2%
  triton_mm_3536 0.6012 ms 69.8%
  triton_mm_3540 0.6168 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.5770 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3686 0.0528 ms 100.0%
  mm 0.0529 ms 99.9%
  triton_mm_3682 0.0538 ms 98.2%
  triton_mm_3690 0.0574 ms 91.9%
  triton_mm_3694 0.0680 ms 77.6%
  triton_mm_3681 0.0848 ms 62.3%
  triton_mm_3680 0.0883 ms 59.8%
  triton_mm_3685 0.0944 ms 55.9%
  triton_mm_3689 0.1009 ms 52.3%
  triton_mm_3679 0.1038 ms 50.9%
SingleProcess AUTOTUNE benchmarking takes 2.1486 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3727 0.4208 ms 100.0%
  triton_mm_3728 0.4276 ms 98.4%
  mm 0.4384 ms 96.0%
  triton_mm_3729 0.4665 ms 90.2%
  triton_mm_3722 0.5306 ms 79.3%
  triton_mm_3720 0.5406 ms 77.8%
  triton_mm_3724 0.5409 ms 77.8%
  triton_mm_3723 0.5802 ms 72.5%
  triton_mm_3721 0.5992 ms 70.2%
  triton_mm_3725 0.6167 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 2.5771 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3871 0.0517 ms 100.0%
  mm 0.0528 ms 98.0%
  triton_mm_3867 0.0533 ms 97.0%
  triton_mm_3875 0.0583 ms 88.7%
  triton_mm_3879 0.0681 ms 76.0%
  triton_mm_3866 0.0849 ms 60.9%
  triton_mm_3865 0.0884 ms 58.5%
  triton_mm_3870 0.0951 ms 54.4%
  triton_mm_3874 0.1009 ms 51.3%
  triton_mm_3864 0.1037 ms 49.9%
SingleProcess AUTOTUNE benchmarking takes 2.1560 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3912 0.4171 ms 100.0%
  triton_mm_3913 0.4259 ms 97.9%
  mm 0.4430 ms 94.1%
  triton_mm_3914 0.4643 ms 89.8%
  triton_mm_3907 0.5304 ms 78.6%
  triton_mm_3909 0.5385 ms 77.5%
  triton_mm_3905 0.5389 ms 77.4%
  triton_mm_3908 0.5840 ms 71.4%
  triton_mm_3906 0.5987 ms 69.7%
  triton_mm_3910 0.6132 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.5787 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0546 ms 100.0%
  triton_mm_7428 0.0570 ms 95.8%
  triton_mm_7420 0.0633 ms 86.3%
  triton_mm_7424 0.0639 ms 85.5%
  triton_mm_7422 0.0650 ms 84.1%
  mm 0.0671 ms 81.4%
  triton_mm_7421 0.0690 ms 79.2%
  triton_mm_7429 0.0696 ms 78.5%
  triton_mm_7425 0.0713 ms 76.6%
  triton_mm_7423 0.0852 ms 64.1%
SingleProcess AUTOTUNE benchmarking takes 12.5791 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1181 ms 100.0%
  triton_mm_7465 0.1209 ms 97.7%
  triton_mm_7466 0.1252 ms 94.3%
  triton_mm_7467 0.1309 ms 90.2%
  triton_mm_7458 0.1523 ms 77.5%
  triton_mm_7460 0.1525 ms 77.4%
  triton_mm_7462 0.1569 ms 75.3%
  triton_mm_7459 0.1643 ms 71.9%
  triton_mm_7461 0.1645 ms 71.8%
  triton_mm_7463 0.1675 ms 70.5%
SingleProcess AUTOTUNE benchmarking takes 10.5344 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1268 ms 100.0%
  triton_mm_7484 0.1452 ms 87.3%
  triton_mm_7485 0.1487 ms 85.3%
  triton_mm_7486 0.1531 ms 82.8%
  triton_mm_7477 0.1611 ms 78.7%
  triton_mm_7481 0.1624 ms 78.1%
  triton_mm_7479 0.1657 ms 76.5%
  triton_mm_7480 0.1668 ms 76.0%
  triton_mm_7482 0.1690 ms 75.0%
  triton_mm_7478 0.1700 ms 74.6%
SingleProcess AUTOTUNE benchmarking takes 10.2279 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1297 ms 100.0%
  triton_mm_7522 0.1431 ms 90.6%
  triton_mm_7523 0.1481 ms 87.6%
  triton_mm_7519 0.1498 ms 86.6%
  triton_mm_7515 0.1512 ms 85.7%
  triton_mm_7524 0.1530 ms 84.7%
  triton_mm_7518 0.1631 ms 79.5%
  triton_mm_7517 0.1662 ms 78.0%
  triton_mm_7520 0.1688 ms 76.8%
  triton_mm_7516 0.1714 ms 75.7%
SingleProcess AUTOTUNE benchmarking takes 10.1690 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0554 ms 100.0%
  triton_mm_7448 0.0560 ms 98.9%
  triton_mm_7447 0.0609 ms 90.9%
  triton_mm_7441 0.0645 ms 85.9%
  triton_mm_7439 0.0699 ms 79.2%
  triton_mm_7446 0.0701 ms 79.0%
  triton_mm_7443 0.0701 ms 78.9%
  triton_mm_7442 0.0711 ms 77.8%
  triton_mm_7444 0.0730 ms 75.8%
  triton_mm_7440 0.0735 ms 75.4%
SingleProcess AUTOTUNE benchmarking takes 10.1543 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1175 ms 100.0%
  triton_mm_7503 0.1204 ms 97.6%
  triton_mm_7504 0.1253 ms 93.8%
  triton_mm_7505 0.1283 ms 91.6%
  triton_mm_7500 0.1476 ms 79.6%
  triton_mm_7496 0.1482 ms 79.3%
  triton_mm_7498 0.1484 ms 79.2%
  triton_mm_7499 0.1568 ms 74.9%
  triton_mm_7497 0.1636 ms 71.8%
  triton_mm_7501 0.1690 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 10.7808 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  triton_mm_8637 0.0338 ms 100.0%
  mm 0.0341 ms 99.2%
  triton_mm_8638 0.0395 ms 85.7%
  triton_mm_8645 0.0404 ms 83.7%
  triton_mm_8636 0.0408 ms 82.9%
  triton_mm_8642 0.0415 ms 81.6%
  triton_mm_8646 0.0425 ms 79.6%
  triton_mm_8639 0.0431 ms 78.4%
  triton_mm_8650 0.0449 ms 75.3%
  triton_mm_8644 0.0461 ms 73.3%
SingleProcess AUTOTUNE benchmarking takes 2.1096 seconds and 0.1326 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_9 0.0247 ms 100.0%
  triton_mm_11 0.0247 ms 100.0%
  triton_mm_6 0.0250 ms 99.0%
  triton_mm_8 0.0250 ms 99.0%
  triton_mm_7 0.0252 ms 97.8%
  triton_mm_4 0.0255 ms 97.0%
  triton_mm_12 0.0260 ms 95.1%
  triton_mm_2 0.0263 ms 94.0%
  triton_mm_3 0.0270 ms 91.4%
  triton_mm_1 0.0277 ms 89.0%
SingleProcess AUTOTUNE benchmarking takes 2.7371 seconds and 0.0131 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0475 ms 100.0%
  triton_mm_33 0.0482 ms 98.5%
  triton_mm_25 0.0485 ms 97.8%
  triton_mm_24 0.0493 ms 96.3%
  triton_mm_28 0.0494 ms 96.1%
  triton_mm_18 0.0501 ms 94.6%
  triton_mm_21 0.0505 ms 93.9%
  triton_mm_19 0.0515 ms 92.2%
  bias_addmm 0.0517 ms 91.8%
  triton_mm_20 0.0540 ms 87.9%
SingleProcess AUTOTUNE benchmarking takes 2.2000 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3412 ms 100.0%
  triton_mm_50 0.3506 ms 97.3%
  triton_mm_52 0.3758 ms 90.8%
  bias_addmm 0.4015 ms 85.0%
  triton_mm_45 0.4058 ms 84.1%
  triton_mm_47 0.4203 ms 81.2%
  triton_mm_43 0.4311 ms 79.1%
  triton_mm_46 0.4483 ms 76.1%
  addmm 0.4643 ms 73.5%
  triton_mm_48 0.4871 ms 70.1%
SingleProcess AUTOTUNE benchmarking takes 2.6806 seconds and 0.0241 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1176 ms 100.0%
  triton_mm_69 0.1199 ms 98.1%
  triton_mm_71 0.1265 ms 93.0%
  bias_addmm 0.1302 ms 90.4%
  triton_mm_64 0.1407 ms 83.6%
  triton_mm_66 0.1467 ms 80.2%
  triton_mm_62 0.1470 ms 80.0%
  addmm 0.1473 ms 79.8%
  triton_mm_65 0.1548 ms 76.0%
  triton_mm_67 0.1674 ms 70.3%
SingleProcess AUTOTUNE benchmarking takes 2.5705 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4369 ms 100.0%
  triton_mm_88 0.4428 ms 98.7%
  triton_mm_90 0.4796 ms 91.1%
  bias_addmm 0.5179 ms 84.4%
  triton_mm_83 0.5357 ms 81.5%
  triton_mm_85 0.5555 ms 78.6%
  triton_mm_81 0.5725 ms 76.3%
  triton_mm_84 0.5898 ms 74.1%
  triton_mm_86 0.6259 ms 69.8%
  triton_mm_82 0.6481 ms 67.4%
SingleProcess AUTOTUNE benchmarking takes 2.7153 seconds and 1.8576 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4270 ms 100.0%
  triton_mm_108 0.4318 ms 98.9%
  triton_mm_107 0.4531 ms 94.2%
  triton_mm_109 0.4542 ms 94.0%
  bias_addmm 0.4947 ms 86.3%
  triton_mm_102 0.5360 ms 79.7%
  triton_mm_100 0.5612 ms 76.1%
  triton_mm_103 0.5616 ms 76.0%
  triton_mm_104 0.5621 ms 76.0%
  triton_mm_105 0.6255 ms 68.3%
SingleProcess AUTOTUNE benchmarking takes 2.6986 seconds and 0.2495 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_122 0.0476 ms 100.0%
  triton_mm_126 0.0482 ms 98.7%
  triton_mm_118 0.0486 ms 97.8%
  triton_mm_117 0.0494 ms 96.3%
  triton_mm_111 0.0495 ms 96.1%
  triton_mm_121 0.0503 ms 94.5%
  triton_mm_112 0.0512 ms 92.9%
  triton_mm_114 0.0514 ms 92.5%
  bias_addmm 0.0518 ms 91.8%
  triton_mm_113 0.0547 ms 87.0%
SingleProcess AUTOTUNE benchmarking takes 2.2025 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4310 ms 100.0%
  addmm 0.4335 ms 99.4%
  triton_mm_200 0.4520 ms 95.4%
  triton_mm_202 0.4548 ms 94.8%
  bias_addmm 0.4937 ms 87.3%
  triton_mm_195 0.5365 ms 80.3%
  triton_mm_193 0.5593 ms 77.1%
  triton_mm_196 0.5604 ms 76.9%
  triton_mm_197 0.5612 ms 76.8%
  triton_mm_198 0.6252 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7035 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_215 0.0467 ms 100.0%
  triton_mm_211 0.0486 ms 96.1%
  triton_mm_219 0.0489 ms 95.5%
  triton_mm_214 0.0493 ms 94.8%
  triton_mm_210 0.0494 ms 94.6%
  triton_mm_204 0.0495 ms 94.4%
  triton_mm_207 0.0514 ms 91.0%
  triton_mm_205 0.0519 ms 90.1%
  bias_addmm 0.0523 ms 89.4%
  triton_mm_206 0.0541 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 2.2021 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4256 ms 100.0%
  triton_mm_294 0.4302 ms 98.9%
  triton_mm_293 0.4511 ms 94.4%
  triton_mm_295 0.4537 ms 93.8%
  bias_addmm 0.4946 ms 86.1%
  triton_mm_288 0.5350 ms 79.6%
  triton_mm_286 0.5592 ms 76.1%
  triton_mm_289 0.5592 ms 76.1%
  triton_mm_290 0.5614 ms 75.8%
  triton_mm_291 0.6257 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 2.6989 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_308 0.0467 ms 100.0%
  triton_mm_304 0.0479 ms 97.6%
  triton_mm_312 0.0481 ms 97.1%
  triton_mm_307 0.0492 ms 95.0%
  triton_mm_303 0.0495 ms 94.4%
  triton_mm_297 0.0500 ms 93.4%
  triton_mm_300 0.0513 ms 91.1%
  triton_mm_298 0.0519 ms 90.0%
  bias_addmm 0.0521 ms 89.7%
  triton_mm_310 0.0544 ms 85.9%
SingleProcess AUTOTUNE benchmarking takes 2.1846 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_387 0.4300 ms 100.0%
  triton_mm_386 0.4506 ms 95.4%
  triton_mm_388 0.4527 ms 95.0%
  addmm 0.4922 ms 87.4%
  bias_addmm 0.4938 ms 87.1%
  triton_mm_381 0.5363 ms 80.2%
  triton_mm_379 0.5561 ms 77.3%
  triton_mm_383 0.5599 ms 76.8%
  triton_mm_382 0.5603 ms 76.7%
  triton_mm_384 0.6222 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.7095 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_401 0.0467 ms 100.0%
  triton_mm_397 0.0479 ms 97.5%
  triton_mm_405 0.0490 ms 95.4%
  triton_mm_400 0.0503 ms 92.9%
  triton_mm_396 0.0503 ms 92.9%
  triton_mm_390 0.0503 ms 92.8%
  triton_mm_393 0.0507 ms 92.2%
  triton_mm_391 0.0513 ms 91.0%
  bias_addmm 0.0521 ms 89.7%
  triton_mm_403 0.0543 ms 86.1%
SingleProcess AUTOTUNE benchmarking takes 2.2157 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_480 0.4286 ms 100.0%
  triton_mm_479 0.4495 ms 95.4%
  triton_mm_481 0.4524 ms 94.8%
  addmm 0.4919 ms 87.1%
  bias_addmm 0.4944 ms 86.7%
  triton_mm_474 0.5337 ms 80.3%
  triton_mm_472 0.5567 ms 77.0%
  triton_mm_476 0.5580 ms 76.8%
  triton_mm_475 0.5587 ms 76.7%
  triton_mm_477 0.6230 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7122 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_494 0.0475 ms 100.0%
  triton_mm_498 0.0482 ms 98.5%
  triton_mm_490 0.0486 ms 97.7%
  triton_mm_493 0.0492 ms 96.4%
  triton_mm_489 0.0493 ms 96.2%
  triton_mm_483 0.0495 ms 95.8%
  triton_mm_486 0.0506 ms 93.8%
  triton_mm_484 0.0513 ms 92.5%
  bias_addmm 0.0524 ms 90.6%
  triton_mm_485 0.0547 ms 86.8%
SingleProcess AUTOTUNE benchmarking takes 2.2224 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1165 ms 100.0%
  triton_mm_591 0.1212 ms 96.2%
  triton_mm_593 0.1270 ms 91.8%
  triton_mm_586 0.1434 ms 81.3%
  triton_mm_588 0.1488 ms 78.3%
  triton_mm_584 0.1499 ms 77.8%
  addmm 0.1508 ms 77.3%
  bias_addmm 0.1536 ms 75.9%
  triton_mm_587 0.1631 ms 71.5%
  triton_mm_589 0.1674 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 11.1302 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1229 ms 100.0%
  triton_mm_610 0.1255 ms 98.0%
  triton_mm_612 0.1339 ms 91.8%
  bias_addmm 0.1352 ms 91.0%
  triton_mm_607 0.1526 ms 80.6%
  addmm 0.1551 ms 79.3%
  triton_mm_605 0.1553 ms 79.2%
  triton_mm_603 0.1560 ms 78.8%
  triton_mm_606 0.1739 ms 70.7%
  triton_mm_608 0.1757 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 2.5768 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0586 ms 100.0%
  triton_mm_624 0.0605 ms 96.8%
  triton_mm_626 0.0609 ms 96.3%
  triton_mm_622 0.0622 ms 94.2%
  triton_mm_630 0.0637 ms 92.0%
  triton_mm_629 0.0637 ms 92.0%
  triton_mm_625 0.0678 ms 86.5%
  triton_mm_631 0.0679 ms 86.4%
  triton_mm_627 0.0711 ms 82.5%
  triton_mm_623 0.0719 ms 81.5%
SingleProcess AUTOTUNE benchmarking takes 2.3794 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_644 0.0467 ms 100.0%
  triton_mm_640 0.0478 ms 97.7%
  triton_mm_648 0.0482 ms 96.9%
  triton_mm_633 0.0496 ms 94.2%
  triton_mm_639 0.0501 ms 93.2%
  triton_mm_643 0.0501 ms 93.2%
  triton_mm_636 0.0505 ms 92.5%
  bias_addmm 0.0517 ms 90.3%
  triton_mm_634 0.0519 ms 90.1%
  triton_mm_635 0.0540 ms 86.5%
SingleProcess AUTOTUNE benchmarking takes 2.2217 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_737 0.0475 ms 100.0%
  triton_mm_733 0.0478 ms 99.3%
  triton_mm_741 0.0482 ms 98.5%
  triton_mm_732 0.0494 ms 96.2%
  triton_mm_726 0.0502 ms 94.7%
  triton_mm_736 0.0503 ms 94.4%
  triton_mm_729 0.0514 ms 92.4%
  triton_mm_727 0.0520 ms 91.3%
  bias_addmm 0.0524 ms 90.7%
  triton_mm_728 0.0540 ms 87.9%
SingleProcess AUTOTUNE benchmarking takes 2.2135 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_830 0.0475 ms 100.0%
  triton_mm_826 0.0478 ms 99.3%
  triton_mm_834 0.0482 ms 98.5%
  triton_mm_819 0.0498 ms 95.4%
  triton_mm_825 0.0501 ms 94.7%
  triton_mm_829 0.0501 ms 94.7%
  triton_mm_822 0.0505 ms 94.1%
  triton_mm_820 0.0519 ms 91.4%
  bias_addmm 0.0523 ms 90.8%
  triton_mm_821 0.0540 ms 87.9%
SingleProcess AUTOTUNE benchmarking takes 2.2074 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_923 0.0476 ms 100.0%
  triton_mm_919 0.0485 ms 98.0%
  triton_mm_927 0.0489 ms 97.2%
  triton_mm_922 0.0492 ms 96.6%
  triton_mm_918 0.0494 ms 96.2%
  triton_mm_912 0.0495 ms 96.0%
  triton_mm_915 0.0513 ms 92.6%
  triton_mm_913 0.0520 ms 91.4%
  bias_addmm 0.0524 ms 90.8%
  triton_mm_914 0.0540 ms 88.0%
SingleProcess AUTOTUNE benchmarking takes 2.9231 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_1016 0.0466 ms 100.0%
  triton_mm_1012 0.0479 ms 97.4%
  triton_mm_1020 0.0482 ms 96.7%
  triton_mm_1015 0.0493 ms 94.5%
  triton_mm_1011 0.0493 ms 94.5%
  triton_mm_1005 0.0497 ms 93.9%
  triton_mm_1008 0.0506 ms 92.2%
  bias_addmm 0.0520 ms 89.7%
  triton_mm_1006 0.0520 ms 89.7%
  triton_mm_1018 0.0544 ms 85.7%
SingleProcess AUTOTUNE benchmarking takes 2.2293 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_1109 0.0467 ms 100.0%
  triton_mm_1105 0.0478 ms 97.6%
  triton_mm_1113 0.0481 ms 97.1%
  triton_mm_1104 0.0493 ms 94.7%
  triton_mm_1098 0.0501 ms 93.2%
  triton_mm_1108 0.0504 ms 92.7%
  triton_mm_1101 0.0505 ms 92.5%
  triton_mm_1099 0.0514 ms 90.8%
  bias_addmm 0.0521 ms 89.6%
  triton_mm_1111 0.0543 ms 86.0%
SingleProcess AUTOTUNE benchmarking takes 2.2242 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0215 ms 100.0%
  triton_mm_2872 0.0228 ms 94.2%
  triton_mm_2880 0.0229 ms 93.9%
  triton_mm_2868 0.0232 ms 92.8%
  bias_addmm 0.0254 ms 84.7%
  triton_mm_2865 0.0270 ms 79.6%
  triton_mm_2871 0.0275 ms 78.1%
  addmm 0.0276 ms 78.0%
  triton_mm_2875 0.0276 ms 77.9%
  triton_mm_2866 0.0280 ms 76.9%
SingleProcess AUTOTUNE benchmarking takes 7.6568 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2897 1.8867 ms 100.0%
  bias_addmm 1.9815 ms 95.2%
  triton_mm_2898 2.0117 ms 93.8%
  triton_mm_2899 2.0527 ms 91.9%
  addmm 2.1851 ms 86.3%
  triton_mm_2892 2.2612 ms 83.4%
  triton_mm_2894 2.3318 ms 80.9%
  triton_mm_2890 2.4092 ms 78.3%
  triton_mm_2893 2.4834 ms 76.0%
  triton_mm_2895 2.6801 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 11.8968 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  mm 1.6988 ms 100.0%
  triton_mm_2935 1.7219 ms 98.7%
  triton_mm_2936 1.7719 ms 95.9%
  triton_mm_2937 1.8963 ms 89.6%
  triton_mm_2928 2.2355 ms 76.0%
  triton_mm_2930 2.2556 ms 75.3%
  triton_mm_2932 2.3138 ms 73.4%
  triton_mm_2931 2.3330 ms 72.8%
  triton_mm_2933 2.4098 ms 70.5%
  triton_mm_2929 2.4426 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 9.9166 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.6882 ms 100.0%
  triton_mm_2916 1.8374 ms 91.9%
  triton_mm_2918 1.8726 ms 90.2%
  triton_mm_2917 1.8800 ms 89.8%
  triton_mm_2912 2.3173 ms 72.9%
  triton_mm_2911 2.4318 ms 69.4%
  triton_mm_2909 2.4384 ms 69.2%
  triton_mm_2913 2.4928 ms 67.7%
  triton_mm_2910 2.5130 ms 67.2%
  triton_mm_2914 2.6098 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 13.1944 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4240 ms 100.0%
  triton_mm_3007 0.4380 ms 96.8%
  triton_mm_3008 0.4625 ms 91.7%
  mm 0.4759 ms 89.1%
  triton_mm_3003 0.5261 ms 80.6%
  triton_mm_2999 0.5380 ms 78.8%
  triton_mm_3001 0.5413 ms 78.3%
  triton_mm_3004 0.5678 ms 74.7%
  triton_mm_3002 0.5713 ms 74.2%
  triton_mm_3000 0.5823 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 9.8573 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2962 0.0163 ms 100.0%
  triton_mm_2961 0.0166 ms 98.3%
  triton_mm_2964 0.0166 ms 98.1%
  triton_mm_2963 0.0168 ms 97.0%
  triton_mm_2960 0.0174 ms 93.7%
  triton_mm_2967 0.0174 ms 93.5%
  triton_mm_2966 0.0174 ms 93.4%
  triton_mm_2965 0.0177 ms 92.0%
  triton_mm_2969 0.0179 ms 91.1%
  triton_mm_2970 0.0188 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 5.2870 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0229 ms 100.0%
  triton_mm_2942 0.0253 ms 90.5%
  triton_mm_2950 0.0257 ms 89.0%
  mm 0.0281 ms 81.3%
  triton_mm_2954 0.0296 ms 77.2%
  triton_mm_2940 0.0344 ms 66.5%
  triton_mm_2941 0.0348 ms 65.8%
  triton_mm_2945 0.0367 ms 62.3%
  triton_mm_2949 0.0389 ms 58.8%
  triton_mm_2939 0.0409 ms 55.9%
SingleProcess AUTOTUNE benchmarking takes 2.0923 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4166 ms 100.0%
  triton_mm_2988 0.4205 ms 99.1%
  mm 0.4408 ms 94.5%
  triton_mm_2982 0.5163 ms 80.7%
  triton_mm_2984 0.5247 ms 79.4%
  triton_mm_2989 0.5247 ms 79.4%
  triton_mm_2980 0.5268 ms 79.1%
  triton_mm_2983 0.5644 ms 73.8%
  triton_mm_2981 0.5853 ms 71.2%
  triton_mm_2985 0.6042 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 10.2556 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4171 ms 100.0%
  triton_mm_3044 0.4182 ms 99.7%
  triton_mm_3045 0.4344 ms 96.0%
  triton_mm_3046 0.4596 ms 90.7%
  triton_mm_3037 0.5171 ms 80.7%
  triton_mm_3039 0.5242 ms 79.6%
  triton_mm_3041 0.5252 ms 79.4%
  triton_mm_3040 0.5563 ms 75.0%
  triton_mm_3042 0.5645 ms 73.9%
  triton_mm_3038 0.5688 ms 73.3%
SingleProcess AUTOTUNE benchmarking takes 9.9854 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3968 ms 100.0%
  triton_mm_3025 0.4251 ms 93.3%
  triton_mm_3026 0.4364 ms 90.9%
  triton_mm_3027 0.4416 ms 89.9%
  triton_mm_3020 0.5183 ms 76.6%
  triton_mm_3018 0.5291 ms 75.0%
  triton_mm_3022 0.5374 ms 73.8%
  triton_mm_3021 0.5407 ms 73.4%
  triton_mm_3019 0.5903 ms 67.2%
  triton_mm_3023 0.6124 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 9.9223 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3194 ms 100.0%
  triton_mm_3120 0.3221 ms 99.1%
  triton_mm_3121 0.3341 ms 95.6%
  triton_mm_3122 0.3543 ms 90.2%
  triton_mm_3113 0.3939 ms 81.1%
  triton_mm_3115 0.3976 ms 80.3%
  triton_mm_3117 0.3988 ms 80.1%
  triton_mm_3116 0.4215 ms 75.8%
  triton_mm_3118 0.4289 ms 74.5%
  triton_mm_3114 0.4329 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 9.9855 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3212 ms 100.0%
  triton_mm_3102 0.3223 ms 99.7%
  mm 0.3269 ms 98.3%
  triton_mm_3103 0.3364 ms 95.5%
  triton_mm_3096 0.3878 ms 82.8%
  triton_mm_3094 0.3987 ms 80.6%
  triton_mm_3098 0.4060 ms 79.1%
  triton_mm_3097 0.4102 ms 78.3%
  triton_mm_3095 0.4426 ms 72.6%
  triton_mm_3099 0.4625 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 12.4259 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1404 ms 100.0%
  triton_mm_3075 0.1468 ms 95.6%
  triton_mm_3079 0.1480 ms 94.9%
  triton_mm_3083 0.1485 ms 94.5%
  triton_mm_3077 0.1486 ms 94.5%
  triton_mm_3084 0.1500 ms 93.6%
  triton_mm_3078 0.1653 ms 84.9%
  triton_mm_3080 0.1682 ms 83.5%
  triton_mm_3076 0.1693 ms 82.9%
  mm 0.1779 ms 78.9%
SingleProcess AUTOTUNE benchmarking takes 9.8406 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1111 ms 100.0%
  triton_mm_3064 0.1136 ms 97.7%
  triton_mm_3065 0.1217 ms 91.3%
  triton_mm_3063 0.1325 ms 83.8%
  triton_mm_3058 0.1347 ms 82.5%
  triton_mm_3060 0.1385 ms 80.2%
  triton_mm_3056 0.1394 ms 79.7%
  triton_mm_3059 0.1473 ms 75.4%
  triton_mm_3057 0.1552 ms 71.6%
  triton_mm_3061 0.1608 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.2018 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3152 0.0321 ms 100.0%
  triton_mm_3145 0.0325 ms 98.7%
  triton_mm_3151 0.0330 ms 97.3%
  triton_mm_3149 0.0335 ms 95.8%
  triton_mm_3146 0.0335 ms 95.7%
  triton_mm_3147 0.0337 ms 95.2%
  triton_mm_3154 0.0341 ms 94.1%
  triton_mm_3150 0.0344 ms 93.2%
  triton_mm_3148 0.0345 ms 92.9%
  triton_mm_3155 0.0364 ms 88.2%
SingleProcess AUTOTUNE benchmarking takes 5.3627 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4206 ms 100.0%
  mm 0.4281 ms 98.2%
  triton_mm_3192 0.4356 ms 96.6%
  triton_mm_3193 0.4599 ms 91.5%
  triton_mm_3188 0.5314 ms 79.2%
  triton_mm_3184 0.5365 ms 78.4%
  triton_mm_3186 0.5425 ms 77.5%
  triton_mm_3189 0.5723 ms 73.5%
  triton_mm_3187 0.5734 ms 73.4%
  triton_mm_3185 0.5841 ms 72.0%
SingleProcess AUTOTUNE benchmarking takes 2.5692 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3127 0.0530 ms 100.0%
  mm 0.0531 ms 99.7%
  triton_mm_3131 0.0537 ms 98.7%
  triton_mm_3135 0.0576 ms 91.9%
  triton_mm_3139 0.0694 ms 76.3%
  triton_mm_3126 0.0845 ms 62.7%
  triton_mm_3125 0.0888 ms 59.7%
  triton_mm_3130 0.0949 ms 55.8%
  triton_mm_3134 0.1012 ms 52.3%
  triton_mm_3124 0.1043 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 6.6294 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4136 ms 100.0%
  triton_mm_3173 0.4209 ms 98.3%
  mm 0.4364 ms 94.8%
  triton_mm_3174 0.4604 ms 89.8%
  triton_mm_3167 0.5211 ms 79.4%
  triton_mm_3169 0.5292 ms 78.2%
  triton_mm_3165 0.5312 ms 77.9%
  triton_mm_3168 0.5712 ms 72.4%
  triton_mm_3166 0.5920 ms 69.9%
  triton_mm_3170 0.6071 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.5730 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4115 ms 100.0%
  triton_mm_3229 0.4206 ms 97.8%
  triton_mm_3230 0.4378 ms 94.0%
  triton_mm_3231 0.4653 ms 88.4%
  triton_mm_3222 0.5188 ms 79.3%
  triton_mm_3224 0.5298 ms 77.7%
  triton_mm_3226 0.5316 ms 77.4%
  triton_mm_3225 0.5673 ms 72.5%
  triton_mm_3227 0.5688 ms 72.3%
  triton_mm_3223 0.5759 ms 71.5%
SingleProcess AUTOTUNE benchmarking takes 2.5767 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3974 ms 100.0%
  triton_mm_3210 0.4288 ms 92.7%
  triton_mm_3211 0.4420 ms 89.9%
  triton_mm_3212 0.4468 ms 88.9%
  triton_mm_3205 0.5244 ms 75.8%
  triton_mm_3203 0.5369 ms 74.0%
  triton_mm_3207 0.5447 ms 73.0%
  triton_mm_3206 0.5493 ms 72.4%
  triton_mm_3204 0.5959 ms 66.7%
  triton_mm_3208 0.6186 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.5859 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3196 ms 100.0%
  triton_mm_3305 0.3261 ms 98.0%
  triton_mm_3306 0.3378 ms 94.6%
  triton_mm_3307 0.3572 ms 89.5%
  triton_mm_3298 0.3993 ms 80.1%
  triton_mm_3300 0.4016 ms 79.6%
  triton_mm_3302 0.4036 ms 79.2%
  triton_mm_3301 0.4264 ms 75.0%
  triton_mm_3303 0.4338 ms 73.7%
  triton_mm_3299 0.4389 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 2.5434 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3213 ms 100.0%
  triton_mm_3287 0.3228 ms 99.5%
  mm 0.3280 ms 98.0%
  triton_mm_3288 0.3380 ms 95.1%
  triton_mm_3281 0.3924 ms 81.9%
  triton_mm_3279 0.4029 ms 79.8%
  triton_mm_3283 0.4085 ms 78.7%
  triton_mm_3282 0.4112 ms 78.1%
  triton_mm_3280 0.4480 ms 71.7%
  triton_mm_3284 0.4643 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.5480 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1404 ms 100.0%
  triton_mm_3260 0.1460 ms 96.1%
  triton_mm_3264 0.1472 ms 95.4%
  triton_mm_3262 0.1477 ms 95.0%
  triton_mm_3268 0.1478 ms 95.0%
  triton_mm_3269 0.1497 ms 93.8%
  triton_mm_3263 0.1653 ms 84.9%
  triton_mm_3265 0.1673 ms 83.9%
  triton_mm_3261 0.1692 ms 83.0%
  mm 0.1790 ms 78.4%
SingleProcess AUTOTUNE benchmarking takes 2.4476 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1100 ms 100.0%
  triton_mm_3248 0.1126 ms 97.6%
  triton_mm_3249 0.1136 ms 96.8%
  triton_mm_3250 0.1208 ms 91.0%
  triton_mm_3243 0.1344 ms 81.8%
  triton_mm_3245 0.1381 ms 79.6%
  triton_mm_3241 0.1396 ms 78.8%
  triton_mm_3244 0.1474 ms 74.6%
  triton_mm_3242 0.1551 ms 70.9%
  triton_mm_3246 0.1605 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.4489 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4168 ms 100.0%
  triton_mm_3414 0.4216 ms 98.9%
  triton_mm_3415 0.4395 ms 94.8%
  triton_mm_3416 0.4646 ms 89.7%
  triton_mm_3407 0.5236 ms 79.6%
  triton_mm_3409 0.5305 ms 78.6%
  triton_mm_3411 0.5328 ms 78.2%
  triton_mm_3410 0.5681 ms 73.4%
  triton_mm_3412 0.5695 ms 73.2%
  triton_mm_3408 0.5749 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.5713 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3986 ms 100.0%
  triton_mm_3395 0.4301 ms 92.7%
  triton_mm_3396 0.4416 ms 90.3%
  triton_mm_3397 0.4481 ms 89.0%
  triton_mm_3390 0.5248 ms 76.0%
  triton_mm_3388 0.5409 ms 73.7%
  triton_mm_3392 0.5447 ms 73.2%
  triton_mm_3391 0.5494 ms 72.6%
  triton_mm_3389 0.5955 ms 66.9%
  triton_mm_3393 0.6191 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.5739 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4169 ms 100.0%
  triton_mm_3599 0.4214 ms 98.9%
  triton_mm_3600 0.4378 ms 95.2%
  triton_mm_3601 0.4643 ms 89.8%
  triton_mm_3592 0.5225 ms 79.8%
  triton_mm_3594 0.5298 ms 78.7%
  triton_mm_3596 0.5321 ms 78.3%
  triton_mm_3595 0.5662 ms 73.6%
  triton_mm_3597 0.5735 ms 72.7%
  triton_mm_3593 0.5764 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 2.5685 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3974 ms 100.0%
  triton_mm_3580 0.4294 ms 92.6%
  triton_mm_3581 0.4429 ms 89.7%
  triton_mm_3582 0.4473 ms 88.8%
  triton_mm_3575 0.5238 ms 75.9%
  triton_mm_3573 0.5379 ms 73.9%
  triton_mm_3577 0.5436 ms 73.1%
  triton_mm_3576 0.5503 ms 72.2%
  triton_mm_3574 0.5961 ms 66.7%
  triton_mm_3578 0.6201 ms 64.1%
SingleProcess AUTOTUNE benchmarking takes 2.5829 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4166 ms 100.0%
  triton_mm_3784 0.4214 ms 98.9%
  triton_mm_3785 0.4391 ms 94.9%
  triton_mm_3786 0.4657 ms 89.5%
  triton_mm_3777 0.5235 ms 79.6%
  triton_mm_3779 0.5302 ms 78.6%
  triton_mm_3781 0.5313 ms 78.4%
  triton_mm_3780 0.5674 ms 73.4%
  triton_mm_3782 0.5740 ms 72.6%
  triton_mm_3778 0.5810 ms 71.7%
SingleProcess AUTOTUNE benchmarking takes 2.5715 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3987 ms 100.0%
  triton_mm_3765 0.4293 ms 92.9%
  triton_mm_3766 0.4428 ms 90.0%
  triton_mm_3767 0.4466 ms 89.3%
  triton_mm_3760 0.5244 ms 76.0%
  triton_mm_3758 0.5399 ms 73.8%
  triton_mm_3762 0.5452 ms 73.1%
  triton_mm_3761 0.5500 ms 72.5%
  triton_mm_3759 0.5955 ms 66.9%
  triton_mm_3763 0.6208 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.5819 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4121 ms 100.0%
  triton_mm_3969 0.4214 ms 97.8%
  triton_mm_3970 0.4388 ms 93.9%
  triton_mm_3971 0.4644 ms 88.7%
  triton_mm_3962 0.5243 ms 78.6%
  triton_mm_3964 0.5304 ms 77.7%
  triton_mm_3966 0.5314 ms 77.6%
  triton_mm_3965 0.5675 ms 72.6%
  triton_mm_3967 0.5733 ms 71.9%
  triton_mm_3963 0.5749 ms 71.7%
SingleProcess AUTOTUNE benchmarking takes 2.5762 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4008 ms 100.0%
  triton_mm_3950 0.4288 ms 93.5%
  triton_mm_3951 0.4422 ms 90.6%
  triton_mm_3952 0.4470 ms 89.6%
  triton_mm_3945 0.5250 ms 76.3%
  triton_mm_3943 0.5350 ms 74.9%
  triton_mm_3947 0.5430 ms 73.8%
  triton_mm_3946 0.5497 ms 72.9%
  triton_mm_3944 0.5954 ms 67.3%
  triton_mm_3948 0.6198 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 2.5833 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0552 ms 100.0%
  mm 0.0555 ms 99.6%
  triton_mm_7428 0.0572 ms 96.6%
  triton_mm_7424 0.0638 ms 86.6%
  triton_mm_7422 0.0659 ms 83.8%
  triton_mm_7421 0.0687 ms 80.4%
  triton_mm_7429 0.0693 ms 79.6%
  triton_mm_7425 0.0713 ms 77.5%
  triton_mm_7420 0.0744 ms 74.3%
  triton_mm_7423 0.0846 ms 65.3%
SingleProcess AUTOTUNE benchmarking takes 11.0081 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1187 ms 100.0%
  triton_mm_7465 0.1216 ms 97.6%
  triton_mm_7466 0.1259 ms 94.2%
  triton_mm_7467 0.1316 ms 90.2%
  triton_mm_7460 0.1523 ms 77.9%
  triton_mm_7458 0.1529 ms 77.6%
  triton_mm_7462 0.1558 ms 76.1%
  triton_mm_7459 0.1635 ms 72.6%
  triton_mm_7461 0.1661 ms 71.4%
  triton_mm_7463 0.1668 ms 71.1%
SingleProcess AUTOTUNE benchmarking takes 10.4643 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1263 ms 100.0%
  triton_mm_7484 0.1444 ms 87.5%
  triton_mm_7485 0.1479 ms 85.4%
  triton_mm_7486 0.1520 ms 83.1%
  triton_mm_7477 0.1605 ms 78.7%
  triton_mm_7481 0.1619 ms 78.0%
  triton_mm_7479 0.1644 ms 76.8%
  triton_mm_7480 0.1668 ms 75.7%
  triton_mm_7482 0.1685 ms 74.9%
  triton_mm_7478 0.1694 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 9.9051 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1290 ms 100.0%
  triton_mm_7522 0.1433 ms 90.0%
  triton_mm_7523 0.1475 ms 87.4%
  triton_mm_7519 0.1490 ms 86.6%
  triton_mm_7515 0.1495 ms 86.3%
  triton_mm_7524 0.1510 ms 85.4%
  triton_mm_7518 0.1632 ms 79.0%
  triton_mm_7517 0.1641 ms 78.6%
  triton_mm_7520 0.1695 ms 76.1%
  triton_mm_7516 0.1707 ms 75.6%
SingleProcess AUTOTUNE benchmarking takes 10.0066 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0548 ms 100.0%
  triton_mm_7448 0.0562 ms 97.4%
  triton_mm_7447 0.0602 ms 91.1%
  triton_mm_7441 0.0627 ms 87.3%
  triton_mm_7446 0.0691 ms 79.3%
  triton_mm_7439 0.0694 ms 78.9%
  triton_mm_7443 0.0697 ms 78.6%
  triton_mm_7442 0.0701 ms 78.1%
  triton_mm_7444 0.0719 ms 76.2%
  triton_mm_7440 0.0731 ms 75.0%
SingleProcess AUTOTUNE benchmarking takes 9.9210 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1179 ms 100.0%
  triton_mm_7503 0.1183 ms 99.6%
  triton_mm_7504 0.1249 ms 94.3%
  triton_mm_7505 0.1278 ms 92.2%
  triton_mm_7500 0.1446 ms 81.5%
  triton_mm_7498 0.1466 ms 80.4%
  triton_mm_7496 0.1473 ms 80.0%
  triton_mm_7499 0.1564 ms 75.3%
  triton_mm_7497 0.1625 ms 72.5%
  triton_mm_7501 0.1680 ms 70.1%
SingleProcess AUTOTUNE benchmarking takes 10.3651 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0302 ms 100.0%
  triton_mm_8637 0.0350 ms 86.4%
  triton_mm_8638 0.0397 ms 76.1%
  triton_mm_8636 0.0403 ms 75.0%
  triton_mm_8645 0.0408 ms 74.2%
  triton_mm_8642 0.0417 ms 72.5%
  triton_mm_8639 0.0424 ms 71.4%
  triton_mm_8646 0.0426 ms 70.9%
  triton_mm_8650 0.0451 ms 67.1%
  triton_mm_8644 0.0456 ms 66.3%
SingleProcess AUTOTUNE benchmarking takes 7.3309 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_7 0.0243 ms 100.0%
  triton_mm_6 0.0246 ms 98.6%
  triton_mm_9 0.0247 ms 98.4%
  triton_mm_11 0.0247 ms 98.3%
  triton_mm_8 0.0259 ms 93.8%
  triton_mm_12 0.0260 ms 93.2%
  triton_mm_4 0.0266 ms 91.2%
  triton_mm_15 0.0268 ms 90.7%
  triton_mm_3 0.0273 ms 88.8%
  triton_mm_2 0.0274 ms 88.8%
SingleProcess AUTOTUNE benchmarking takes 2.1482 seconds and 0.0143 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0464 ms 100.0%
  triton_mm_25 0.0476 ms 97.5%
  triton_mm_33 0.0478 ms 97.1%
  triton_mm_28 0.0489 ms 94.9%
  triton_mm_18 0.0494 ms 93.8%
  triton_mm_24 0.0500 ms 92.8%
  triton_mm_19 0.0505 ms 91.9%
  triton_mm_21 0.0511 ms 90.7%
  bias_addmm 0.0515 ms 90.1%
  triton_mm_31 0.0537 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 2.1963 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3530 ms 100.0%
  triton_mm_50 0.3595 ms 98.2%
  triton_mm_52 0.3856 ms 91.6%
  bias_addmm 0.4019 ms 87.8%
  triton_mm_45 0.4152 ms 85.0%
  triton_mm_47 0.4330 ms 81.5%
  triton_mm_43 0.4417 ms 79.9%
  triton_mm_46 0.4592 ms 76.9%
  addmm 0.4628 ms 76.3%
  triton_mm_48 0.4947 ms 71.4%
SingleProcess AUTOTUNE benchmarking takes 2.6648 seconds and 0.0139 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1192 ms 100.0%
  triton_mm_69 0.1230 ms 96.9%
  triton_mm_71 0.1292 ms 92.3%
  bias_addmm 0.1331 ms 89.5%
  triton_mm_64 0.1434 ms 83.1%
  triton_mm_62 0.1498 ms 79.6%
  triton_mm_66 0.1502 ms 79.3%
  addmm 0.1514 ms 78.7%
  triton_mm_65 0.1574 ms 75.7%
  triton_mm_67 0.1710 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.5519 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4446 ms 100.0%
  triton_mm_88 0.4515 ms 98.5%
  triton_mm_90 0.4892 ms 90.9%
  bias_addmm 0.5182 ms 85.8%
  triton_mm_83 0.5455 ms 81.5%
  triton_mm_85 0.5666 ms 78.5%
  triton_mm_81 0.5824 ms 76.3%
  triton_mm_84 0.6036 ms 73.7%
  triton_mm_86 0.6409 ms 69.4%
  triton_mm_82 0.6529 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.6994 seconds and 1.6992 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4350 ms 100.0%
  triton_mm_108 0.4387 ms 99.2%
  triton_mm_107 0.4597 ms 94.6%
  triton_mm_109 0.4638 ms 93.8%
  bias_addmm 0.4970 ms 87.5%
  triton_mm_102 0.5450 ms 79.8%
  triton_mm_103 0.5684 ms 76.5%
  triton_mm_100 0.5702 ms 76.3%
  triton_mm_104 0.5702 ms 76.3%
  triton_mm_105 0.6384 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.6969 seconds and 0.2360 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4393 ms 100.0%
  addmm 0.4461 ms 98.5%
  triton_mm_200 0.4576 ms 96.0%
  triton_mm_202 0.4626 ms 95.0%
  bias_addmm 0.4971 ms 88.4%
  triton_mm_195 0.5435 ms 80.8%
  triton_mm_196 0.5667 ms 77.5%
  triton_mm_193 0.5683 ms 77.3%
  triton_mm_197 0.5699 ms 77.1%
  triton_mm_198 0.6377 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7072 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4342 ms 100.0%
  triton_mm_294 0.4369 ms 99.4%
  triton_mm_293 0.4587 ms 94.7%
  triton_mm_295 0.4612 ms 94.2%
  bias_addmm 0.4978 ms 87.2%
  triton_mm_288 0.5418 ms 80.1%
  triton_mm_289 0.5665 ms 76.7%
  triton_mm_286 0.5670 ms 76.6%
  triton_mm_290 0.5680 ms 76.4%
  triton_mm_291 0.6362 ms 68.3%
SingleProcess AUTOTUNE benchmarking takes 2.6866 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_387 0.4359 ms 100.0%
  triton_mm_386 0.4572 ms 95.4%
  triton_mm_388 0.4608 ms 94.6%
  bias_addmm 0.4941 ms 88.2%
  addmm 0.4956 ms 88.0%
  triton_mm_381 0.5400 ms 80.7%
  triton_mm_379 0.5646 ms 77.2%
  triton_mm_382 0.5654 ms 77.1%
  triton_mm_383 0.5667 ms 76.9%
  triton_mm_384 0.6348 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7073 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_480 0.4351 ms 100.0%
  triton_mm_479 0.4560 ms 95.4%
  triton_mm_481 0.4631 ms 93.9%
  addmm 0.4916 ms 88.5%
  bias_addmm 0.4945 ms 88.0%
  triton_mm_474 0.5398 ms 80.6%
  triton_mm_472 0.5651 ms 77.0%
  triton_mm_476 0.5657 ms 76.9%
  triton_mm_475 0.5669 ms 76.7%
  triton_mm_477 0.6342 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.7098 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  addmm 0.4312 ms 100.0%
  triton_mm_573 0.4369 ms 98.7%
  triton_mm_572 0.4564 ms 94.5%
  triton_mm_574 0.4605 ms 93.6%
  bias_addmm 0.4939 ms 87.3%
  triton_mm_567 0.5384 ms 80.1%
  triton_mm_568 0.5644 ms 76.4%
  triton_mm_565 0.5652 ms 76.3%
  triton_mm_569 0.5671 ms 76.0%
  triton_mm_570 0.6349 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 2.6945 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1178 ms 100.0%
  triton_mm_591 0.1209 ms 97.4%
  triton_mm_593 0.1281 ms 92.0%
  triton_mm_586 0.1447 ms 81.4%
  triton_mm_588 0.1489 ms 79.1%
  triton_mm_584 0.1503 ms 78.4%
  addmm 0.1517 ms 77.6%
  bias_addmm 0.1539 ms 76.5%
  triton_mm_587 0.1630 ms 72.2%
  triton_mm_589 0.1685 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.5704 seconds and 0.0057 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1242 ms 100.0%
  triton_mm_610 0.1263 ms 98.3%
  triton_mm_612 0.1346 ms 92.2%
  bias_addmm 0.1348 ms 92.1%
  triton_mm_607 0.1537 ms 80.8%
  triton_mm_605 0.1553 ms 79.9%
  addmm 0.1558 ms 79.7%
  triton_mm_603 0.1572 ms 79.0%
  triton_mm_606 0.1729 ms 71.8%
  triton_mm_608 0.1776 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.5689 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0585 ms 100.0%
  triton_mm_624 0.0596 ms 98.2%
  triton_mm_622 0.0610 ms 96.0%
  triton_mm_626 0.0614 ms 95.4%
  triton_mm_629 0.0630 ms 92.8%
  triton_mm_630 0.0636 ms 92.1%
  triton_mm_631 0.0678 ms 86.3%
  triton_mm_625 0.0688 ms 85.1%
  triton_mm_627 0.0714 ms 82.0%
  triton_mm_623 0.0733 ms 79.8%
SingleProcess AUTOTUNE benchmarking takes 2.3762 seconds and 0.0013 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_723 0.4343 ms 100.0%
  triton_mm_722 0.4537 ms 95.7%
  triton_mm_724 0.4571 ms 95.0%
  addmm 0.4927 ms 88.2%
  bias_addmm 0.4949 ms 87.7%
  triton_mm_717 0.5371 ms 80.9%
  triton_mm_718 0.5603 ms 77.5%
  triton_mm_715 0.5607 ms 77.5%
  triton_mm_719 0.5625 ms 77.2%
  triton_mm_720 0.6296 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7091 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_816 0.4338 ms 100.0%
  addmm 0.4349 ms 99.7%
  triton_mm_815 0.4543 ms 95.5%
  triton_mm_817 0.4572 ms 94.9%
  bias_addmm 0.4952 ms 87.6%
  triton_mm_810 0.5373 ms 80.7%
  triton_mm_811 0.5600 ms 77.5%
  triton_mm_808 0.5608 ms 77.4%
  triton_mm_812 0.5634 ms 77.0%
  triton_mm_813 0.6305 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.6980 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_909 0.4324 ms 100.0%
  triton_mm_908 0.4530 ms 95.5%
  triton_mm_910 0.4595 ms 94.1%
  addmm 0.4927 ms 87.8%
  bias_addmm 0.4952 ms 87.3%
  triton_mm_903 0.5367 ms 80.6%
  triton_mm_904 0.5605 ms 77.2%
  triton_mm_901 0.5606 ms 77.1%
  triton_mm_905 0.5627 ms 76.9%
  triton_mm_906 0.6297 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7131 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_1002 0.4336 ms 100.0%
  addmm 0.4355 ms 99.6%
  triton_mm_1001 0.4534 ms 95.6%
  triton_mm_1003 0.4581 ms 94.6%
  bias_addmm 0.4956 ms 87.5%
  triton_mm_996 0.5366 ms 80.8%
  triton_mm_997 0.5614 ms 77.2%
  triton_mm_994 0.5631 ms 77.0%
  triton_mm_998 0.5635 ms 76.9%
  triton_mm_999 0.6307 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7107 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_1095 0.4338 ms 100.0%
  addmm 0.4357 ms 99.6%
  triton_mm_1094 0.4540 ms 95.5%
  triton_mm_1096 0.4572 ms 94.9%
  bias_addmm 0.4955 ms 87.6%
  triton_mm_1089 0.5377 ms 80.7%
  triton_mm_1090 0.5608 ms 77.4%
  triton_mm_1087 0.5612 ms 77.3%
  triton_mm_1091 0.5627 ms 77.1%
  triton_mm_1092 0.6298 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7064 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0214 ms 100.0%
  triton_mm_2872 0.0227 ms 94.6%
  triton_mm_2880 0.0227 ms 94.5%
  triton_mm_2868 0.0235 ms 91.4%
  bias_addmm 0.0263 ms 81.4%
  triton_mm_2871 0.0274 ms 78.3%
  triton_mm_2875 0.0274 ms 78.3%
  triton_mm_2865 0.0278 ms 77.2%
  triton_mm_2866 0.0278 ms 77.2%
  triton_mm_2867 0.0283 ms 75.8%
SingleProcess AUTOTUNE benchmarking takes 7.6704 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2898 1.8716 ms 100.0%
  triton_mm_2897 1.8980 ms 98.6%
  bias_addmm 1.9840 ms 94.3%
  triton_mm_2899 2.0641 ms 90.7%
  triton_mm_2892 2.2929 ms 81.6%
  triton_mm_2894 2.3646 ms 79.2%
  triton_mm_2890 2.4206 ms 77.3%
  addmm 2.4658 ms 75.9%
  triton_mm_2893 2.4892 ms 75.2%
  triton_mm_2895 2.6930 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 11.9214 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  triton_mm_2935 1.7482 ms 100.0%
  triton_mm_2936 1.8160 ms 96.3%
  mm 1.8789 ms 93.0%
  triton_mm_2937 1.9565 ms 89.4%
  triton_mm_2928 2.2826 ms 76.6%
  triton_mm_2930 2.3039 ms 75.9%
  triton_mm_2932 2.3674 ms 73.8%
  triton_mm_2931 2.4000 ms 72.8%
  triton_mm_2933 2.4725 ms 70.7%
  triton_mm_2929 2.4752 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.8475 seconds and 0.1790 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.7207 ms 100.0%
  triton_mm_2916 1.8847 ms 91.3%
  triton_mm_2917 1.9198 ms 89.6%
  triton_mm_2918 1.9207 ms 89.6%
  triton_mm_2912 2.3427 ms 73.4%
  triton_mm_2911 2.4729 ms 69.6%
  triton_mm_2909 2.5008 ms 68.8%
  triton_mm_2910 2.5607 ms 67.2%
  triton_mm_2913 2.6304 ms 65.4%
  triton_mm_2914 2.6556 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 9.8446 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4285 ms 100.0%
  triton_mm_3007 0.4421 ms 96.9%
  triton_mm_3008 0.4676 ms 91.6%
  mm 0.4783 ms 89.6%
  triton_mm_3003 0.5308 ms 80.7%
  triton_mm_2999 0.5405 ms 79.3%
  triton_mm_3001 0.5455 ms 78.5%
  triton_mm_3004 0.5753 ms 74.5%
  triton_mm_3002 0.5794 ms 73.9%
  triton_mm_3000 0.5907 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 9.8719 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2962 0.0162 ms 100.0%
  triton_mm_2964 0.0164 ms 98.4%
  triton_mm_2967 0.0164 ms 98.4%
  triton_mm_2965 0.0167 ms 97.1%
  triton_mm_2961 0.0171 ms 94.9%
  triton_mm_2960 0.0171 ms 94.6%
  triton_mm_2966 0.0174 ms 92.9%
  triton_mm_2963 0.0176 ms 92.2%
  triton_mm_2969 0.0178 ms 90.8%
  triton_mm_2957 0.0183 ms 88.6%
SingleProcess AUTOTUNE benchmarking takes 2.0181 seconds and 0.0007 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0228 ms 100.0%
  triton_mm_2942 0.0252 ms 90.7%
  triton_mm_2950 0.0256 ms 89.3%
  mm 0.0280 ms 81.7%
  triton_mm_2954 0.0296 ms 77.1%
  triton_mm_2940 0.0350 ms 65.3%
  triton_mm_2941 0.0353 ms 64.7%
  triton_mm_2945 0.0373 ms 61.3%
  triton_mm_2949 0.0388 ms 59.0%
  triton_mm_2939 0.0414 ms 55.2%
SingleProcess AUTOTUNE benchmarking takes 8.9859 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4221 ms 100.0%
  triton_mm_2988 0.4275 ms 98.7%
  mm 0.4476 ms 94.3%
  triton_mm_2989 0.4656 ms 90.7%
  triton_mm_2982 0.5297 ms 79.7%
  triton_mm_2984 0.5349 ms 78.9%
  triton_mm_2980 0.5387 ms 78.4%
  triton_mm_2983 0.5771 ms 73.1%
  triton_mm_2981 0.5992 ms 70.4%
  triton_mm_2985 0.6135 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 10.2191 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4226 ms 100.0%
  triton_mm_3044 0.4278 ms 98.8%
  triton_mm_3045 0.4447 ms 95.0%
  triton_mm_3046 0.4707 ms 89.8%
  triton_mm_3037 0.5268 ms 80.2%
  triton_mm_3041 0.5340 ms 79.1%
  triton_mm_3039 0.5365 ms 78.8%
  triton_mm_3040 0.5666 ms 74.6%
  triton_mm_3042 0.5785 ms 73.1%
  triton_mm_3038 0.5827 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 9.9815 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4076 ms 100.0%
  triton_mm_3025 0.4304 ms 94.7%
  triton_mm_3026 0.4426 ms 92.1%
  triton_mm_3027 0.5031 ms 81.0%
  triton_mm_3020 0.5236 ms 77.8%
  triton_mm_3018 0.5354 ms 76.1%
  triton_mm_3022 0.5443 ms 74.9%
  triton_mm_3021 0.5471 ms 74.5%
  triton_mm_3019 0.5978 ms 68.2%
  triton_mm_3023 0.6201 ms 65.7%
SingleProcess AUTOTUNE benchmarking takes 9.9709 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3238 ms 100.0%
  triton_mm_3120 0.3303 ms 98.1%
  triton_mm_3121 0.3429 ms 94.4%
  triton_mm_3113 0.4033 ms 80.3%
  triton_mm_3115 0.4086 ms 79.3%
  triton_mm_3117 0.4087 ms 79.2%
  triton_mm_3122 0.4098 ms 79.0%
  triton_mm_3116 0.4311 ms 75.1%
  triton_mm_3118 0.4411 ms 73.4%
  triton_mm_3114 0.4438 ms 73.0%
SingleProcess AUTOTUNE benchmarking takes 11.3738 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3258 ms 100.0%
  triton_mm_3102 0.3278 ms 99.4%
  mm 0.3300 ms 98.7%
  triton_mm_3103 0.3854 ms 84.5%
  triton_mm_3096 0.3959 ms 82.3%
  triton_mm_3094 0.4065 ms 80.2%
  triton_mm_3098 0.4114 ms 79.2%
  triton_mm_3097 0.4143 ms 78.6%
  triton_mm_3095 0.4538 ms 71.8%
  triton_mm_3099 0.4703 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 9.8851 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1420 ms 100.0%
  triton_mm_3075 0.1487 ms 95.5%
  triton_mm_3083 0.1487 ms 95.5%
  triton_mm_3079 0.1489 ms 95.4%
  triton_mm_3077 0.1491 ms 95.3%
  triton_mm_3084 0.1518 ms 93.6%
  triton_mm_3078 0.1661 ms 85.5%
  triton_mm_3080 0.1691 ms 84.0%
  triton_mm_3076 0.1712 ms 83.0%
  mm 0.1790 ms 79.3%
SingleProcess AUTOTUNE benchmarking takes 9.8647 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1118 ms 100.0%
  triton_mm_3063 0.1118 ms 99.9%
  triton_mm_3064 0.1142 ms 97.8%
  triton_mm_3065 0.1219 ms 91.7%
  triton_mm_3058 0.1342 ms 83.3%
  triton_mm_3060 0.1377 ms 81.2%
  triton_mm_3056 0.1395 ms 80.1%
  triton_mm_3059 0.1492 ms 74.9%
  triton_mm_3057 0.1553 ms 72.0%
  triton_mm_3061 0.1619 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.2271 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3145 0.0322 ms 100.0%
  triton_mm_3147 0.0324 ms 99.6%
  triton_mm_3148 0.0332 ms 97.2%
  triton_mm_3150 0.0332 ms 96.9%
  triton_mm_3149 0.0333 ms 96.7%
  triton_mm_3152 0.0333 ms 96.7%
  triton_mm_3151 0.0337 ms 95.7%
  triton_mm_3154 0.0346 ms 93.2%
  triton_mm_3146 0.0347 ms 92.9%
  triton_mm_3153 0.0352 ms 91.5%
SingleProcess AUTOTUNE benchmarking takes 5.3463 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4280 ms 100.0%
  mm 0.4317 ms 99.1%
  triton_mm_3192 0.4440 ms 96.4%
  triton_mm_3193 0.4697 ms 91.1%
  triton_mm_3188 0.5328 ms 80.3%
  triton_mm_3184 0.5430 ms 78.8%
  triton_mm_3186 0.5498 ms 77.9%
  triton_mm_3189 0.5769 ms 74.2%
  triton_mm_3187 0.5823 ms 73.5%
  triton_mm_3185 0.5932 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 2.5730 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3131 0.0520 ms 100.0%
  mm 0.0530 ms 98.2%
  triton_mm_3127 0.0536 ms 97.0%
  triton_mm_3135 0.0585 ms 88.8%
  triton_mm_3139 0.0682 ms 76.3%
  triton_mm_3126 0.0848 ms 61.3%
  triton_mm_3125 0.0884 ms 58.8%
  triton_mm_3130 0.0954 ms 54.5%
  triton_mm_3134 0.1014 ms 51.3%
  triton_mm_3124 0.1041 ms 50.0%
SingleProcess AUTOTUNE benchmarking takes 6.5666 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4212 ms 100.0%
  triton_mm_3173 0.4270 ms 98.6%
  mm 0.4485 ms 93.9%
  triton_mm_3174 0.4683 ms 89.9%
  triton_mm_3167 0.5301 ms 79.5%
  triton_mm_3169 0.5356 ms 78.6%
  triton_mm_3165 0.5389 ms 78.2%
  triton_mm_3168 0.5797 ms 72.7%
  triton_mm_3166 0.6004 ms 70.2%
  triton_mm_3170 0.6135 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.5809 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4227 ms 100.0%
  triton_mm_3229 0.4277 ms 98.8%
  triton_mm_3230 0.4444 ms 95.1%
  triton_mm_3231 0.4713 ms 89.7%
  triton_mm_3222 0.5268 ms 80.2%
  triton_mm_3226 0.5349 ms 79.0%
  triton_mm_3224 0.5373 ms 78.7%
  triton_mm_3225 0.5686 ms 74.3%
  triton_mm_3227 0.5779 ms 73.1%
  triton_mm_3223 0.5832 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.5728 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4052 ms 100.0%
  triton_mm_3210 0.4355 ms 93.0%
  triton_mm_3211 0.4472 ms 90.6%
  triton_mm_3212 0.4545 ms 89.2%
  triton_mm_3205 0.5290 ms 76.6%
  triton_mm_3203 0.5433 ms 74.6%
  triton_mm_3207 0.5499 ms 73.7%
  triton_mm_3206 0.5535 ms 73.2%
  triton_mm_3204 0.6029 ms 67.2%
  triton_mm_3208 0.6264 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 2.5717 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3276 ms 100.0%
  triton_mm_3305 0.3292 ms 99.5%
  triton_mm_3306 0.3428 ms 95.6%
  triton_mm_3307 0.3635 ms 90.1%
  triton_mm_3298 0.4030 ms 81.3%
  triton_mm_3300 0.4087 ms 80.2%
  triton_mm_3302 0.4091 ms 80.1%
  triton_mm_3301 0.4312 ms 76.0%
  triton_mm_3303 0.4410 ms 74.3%
  triton_mm_3299 0.4452 ms 73.6%
SingleProcess AUTOTUNE benchmarking takes 2.5407 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3256 ms 100.0%
  triton_mm_3287 0.3276 ms 99.4%
  mm 0.3318 ms 98.1%
  triton_mm_3288 0.3429 ms 95.0%
  triton_mm_3281 0.3966 ms 82.1%
  triton_mm_3279 0.4086 ms 79.7%
  triton_mm_3283 0.4118 ms 79.1%
  triton_mm_3282 0.4151 ms 78.4%
  triton_mm_3280 0.4541 ms 71.7%
  triton_mm_3284 0.4713 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.5531 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1411 ms 100.0%
  triton_mm_3260 0.1478 ms 95.4%
  triton_mm_3264 0.1479 ms 95.4%
  triton_mm_3268 0.1486 ms 94.9%
  triton_mm_3262 0.1489 ms 94.7%
  triton_mm_3269 0.1508 ms 93.5%
  triton_mm_3263 0.1652 ms 85.4%
  triton_mm_3265 0.1688 ms 83.6%
  triton_mm_3261 0.1708 ms 82.6%
  mm 0.1793 ms 78.7%
SingleProcess AUTOTUNE benchmarking takes 2.4426 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1111 ms 100.0%
  triton_mm_3248 0.1132 ms 98.2%
  triton_mm_3249 0.1152 ms 96.4%
  triton_mm_3250 0.1229 ms 90.3%
  triton_mm_3243 0.1363 ms 81.5%
  triton_mm_3245 0.1398 ms 79.4%
  triton_mm_3241 0.1400 ms 79.3%
  triton_mm_3244 0.1493 ms 74.4%
  triton_mm_3242 0.1565 ms 71.0%
  triton_mm_3246 0.1628 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 2.4427 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3334 0.0322 ms 100.0%
  triton_mm_3330 0.0324 ms 99.3%
  triton_mm_3336 0.0326 ms 98.6%
  triton_mm_3335 0.0332 ms 97.0%
  triton_mm_3337 0.0334 ms 96.2%
  triton_mm_3332 0.0336 ms 95.6%
  triton_mm_3333 0.0343 ms 93.7%
  triton_mm_3339 0.0348 ms 92.4%
  triton_mm_3331 0.0349 ms 92.2%
  triton_mm_3338 0.0352 ms 91.4%
SingleProcess AUTOTUNE benchmarking takes 2.0057 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3357 0.4214 ms 100.0%
  triton_mm_3358 0.4267 ms 98.7%
  mm 0.4440 ms 94.9%
  triton_mm_3359 0.4647 ms 90.7%
  triton_mm_3352 0.5299 ms 79.5%
  triton_mm_3354 0.5360 ms 78.6%
  triton_mm_3350 0.5394 ms 78.1%
  triton_mm_3353 0.5804 ms 72.6%
  triton_mm_3351 0.5979 ms 70.5%
  triton_mm_3355 0.6154 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.5768 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4222 ms 100.0%
  triton_mm_3414 0.4254 ms 99.2%
  triton_mm_3415 0.4450 ms 94.9%
  triton_mm_3416 0.4707 ms 89.7%
  triton_mm_3407 0.5260 ms 80.3%
  triton_mm_3411 0.5348 ms 78.9%
  triton_mm_3409 0.5361 ms 78.8%
  triton_mm_3410 0.5691 ms 74.2%
  triton_mm_3412 0.5773 ms 73.1%
  triton_mm_3408 0.5835 ms 72.4%
SingleProcess AUTOTUNE benchmarking takes 2.5787 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4053 ms 100.0%
  triton_mm_3395 0.4348 ms 93.2%
  triton_mm_3396 0.4480 ms 90.5%
  triton_mm_3397 0.4525 ms 89.6%
  triton_mm_3390 0.5283 ms 76.7%
  triton_mm_3388 0.5426 ms 74.7%
  triton_mm_3392 0.5493 ms 73.8%
  triton_mm_3391 0.5533 ms 73.2%
  triton_mm_3389 0.6020 ms 67.3%
  triton_mm_3393 0.6266 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 2.5857 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3542 0.4224 ms 100.0%
  triton_mm_3543 0.4291 ms 98.5%
  mm 0.4465 ms 94.6%
  triton_mm_3544 0.4657 ms 90.7%
  triton_mm_3537 0.5293 ms 79.8%
  triton_mm_3539 0.5363 ms 78.8%
  triton_mm_3535 0.5387 ms 78.4%
  triton_mm_3538 0.5782 ms 73.1%
  triton_mm_3536 0.6005 ms 70.4%
  triton_mm_3540 0.6111 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.5735 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4219 ms 100.0%
  triton_mm_3599 0.4264 ms 98.9%
  triton_mm_3600 0.4438 ms 95.1%
  triton_mm_3601 0.4716 ms 89.5%
  triton_mm_3592 0.5265 ms 80.1%
  triton_mm_3596 0.5335 ms 79.1%
  triton_mm_3594 0.5368 ms 78.6%
  triton_mm_3595 0.5681 ms 74.3%
  triton_mm_3597 0.5782 ms 73.0%
  triton_mm_3593 0.5806 ms 72.7%
SingleProcess AUTOTUNE benchmarking takes 2.5748 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4059 ms 100.0%
  triton_mm_3580 0.4345 ms 93.4%
  triton_mm_3581 0.4480 ms 90.6%
  triton_mm_3582 0.4536 ms 89.5%
  triton_mm_3575 0.5296 ms 76.7%
  triton_mm_3573 0.5426 ms 74.8%
  triton_mm_3577 0.5490 ms 73.9%
  triton_mm_3576 0.5532 ms 73.4%
  triton_mm_3574 0.6035 ms 67.3%
  triton_mm_3578 0.6257 ms 64.9%
SingleProcess AUTOTUNE benchmarking takes 2.5839 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  triton_mm_7427 0.0544 ms 100.0%
  triton_mm_7428 0.0573 ms 95.0%
  triton_mm_7424 0.0637 ms 85.4%
  triton_mm_7420 0.0641 ms 84.9%
  triton_mm_7422 0.0657 ms 82.9%
  mm 0.0663 ms 82.1%
  triton_mm_7421 0.0687 ms 79.2%
  triton_mm_7429 0.0696 ms 78.2%
  triton_mm_7425 0.0714 ms 76.3%
  triton_mm_7423 0.0850 ms 64.0%
SingleProcess AUTOTUNE benchmarking takes 10.9531 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1189 ms 100.0%
  triton_mm_7465 0.1222 ms 97.4%
  triton_mm_7466 0.1275 ms 93.3%
  triton_mm_7467 0.1328 ms 89.6%
  triton_mm_7462 0.1505 ms 79.0%
  triton_mm_7458 0.1530 ms 77.7%
  triton_mm_7460 0.1541 ms 77.2%
  triton_mm_7459 0.1660 ms 71.7%
  triton_mm_7461 0.1670 ms 71.2%
  triton_mm_7463 0.1691 ms 70.3%
SingleProcess AUTOTUNE benchmarking takes 10.5156 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1275 ms 100.0%
  triton_mm_7484 0.1438 ms 88.7%
  triton_mm_7485 0.1495 ms 85.3%
  triton_mm_7481 0.1512 ms 84.3%
  triton_mm_7477 0.1616 ms 78.9%
  triton_mm_7479 0.1646 ms 77.4%
  triton_mm_7480 0.1685 ms 75.6%
  triton_mm_7482 0.1698 ms 75.1%
  triton_mm_7478 0.1716 ms 74.3%
  triton_mm_7486 0.1832 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 9.9291 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1298 ms 100.0%
  triton_mm_7522 0.1437 ms 90.3%
  triton_mm_7523 0.1481 ms 87.6%
  triton_mm_7519 0.1499 ms 86.6%
  triton_mm_7515 0.1501 ms 86.5%
  triton_mm_7524 0.1518 ms 85.5%
  triton_mm_7517 0.1538 ms 84.4%
  triton_mm_7518 0.1642 ms 79.1%
  triton_mm_7520 0.1694 ms 76.6%
  triton_mm_7516 0.1713 ms 75.8%
SingleProcess AUTOTUNE benchmarking takes 9.9815 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0553 ms 100.0%
  triton_mm_7448 0.0558 ms 99.1%
  triton_mm_7447 0.0608 ms 91.0%
  triton_mm_7441 0.0639 ms 86.6%
  triton_mm_7439 0.0687 ms 80.5%
  triton_mm_7446 0.0690 ms 80.2%
  triton_mm_7442 0.0703 ms 78.7%
  triton_mm_7443 0.0705 ms 78.5%
  triton_mm_7444 0.0723 ms 76.6%
  triton_mm_7440 0.0725 ms 76.3%
SingleProcess AUTOTUNE benchmarking takes 11.8644 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  mm 0.1179 ms 100.0%
  triton_mm_7503 0.1204 ms 97.9%
  triton_mm_7504 0.1257 ms 93.8%
  triton_mm_7505 0.1297 ms 90.9%
  triton_mm_7500 0.1444 ms 81.6%
  triton_mm_7498 0.1484 ms 79.5%
  triton_mm_7496 0.1486 ms 79.4%
  triton_mm_7499 0.1567 ms 75.3%
  triton_mm_7497 0.1655 ms 71.2%
  triton_mm_7501 0.1706 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.4966 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0301 ms 100.0%
  triton_mm_8637 0.0347 ms 86.8%
  triton_mm_8638 0.0386 ms 78.0%
  triton_mm_8636 0.0402 ms 74.9%
  triton_mm_8645 0.0411 ms 73.3%
  triton_mm_8642 0.0415 ms 72.6%
  triton_mm_8646 0.0415 ms 72.5%
  triton_mm_8639 0.0424 ms 71.0%
  triton_mm_8650 0.0448 ms 67.3%
  triton_mm_8644 0.0452 ms 66.7%
SingleProcess AUTOTUNE benchmarking takes 7.2130 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_9 0.0247 ms 100.0%
  triton_mm_6 0.0249 ms 99.4%
  triton_mm_8 0.0249 ms 99.4%
  triton_mm_7 0.0253 ms 97.6%
  triton_mm_11 0.0257 ms 96.3%
  triton_mm_12 0.0259 ms 95.6%
  triton_mm_3 0.0265 ms 93.4%
  triton_mm_4 0.0267 ms 92.7%
  triton_mm_2 0.0275 ms 90.0%
  triton_mm_15 0.0278 ms 88.9%
SingleProcess AUTOTUNE benchmarking takes 2.1667 seconds and 0.0098 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_29 0.0466 ms 100.0%
  triton_mm_25 0.0484 ms 96.2%
  triton_mm_33 0.0486 ms 95.7%
  triton_mm_28 0.0498 ms 93.4%
  triton_mm_24 0.0500 ms 93.1%
  triton_mm_18 0.0503 ms 92.6%
  triton_mm_19 0.0506 ms 92.0%
  triton_mm_21 0.0506 ms 92.0%
  bias_addmm 0.0507 ms 91.9%
  triton_mm_31 0.0535 ms 87.0%
SingleProcess AUTOTUNE benchmarking takes 2.1959 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_51 0.3369 ms 100.0%
  triton_mm_50 0.3460 ms 97.4%
  triton_mm_52 0.3708 ms 90.9%
  triton_mm_45 0.3988 ms 84.5%
  bias_addmm 0.3998 ms 84.3%
  triton_mm_47 0.4162 ms 81.0%
  triton_mm_43 0.4256 ms 79.2%
  triton_mm_46 0.4437 ms 75.9%
  addmm 0.4609 ms 73.1%
  triton_mm_48 0.4754 ms 70.9%
SingleProcess AUTOTUNE benchmarking takes 2.6739 seconds and 0.0210 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_70 0.1159 ms 100.0%
  triton_mm_69 0.1188 ms 97.6%
  triton_mm_71 0.1253 ms 92.5%
  bias_addmm 0.1283 ms 90.3%
  triton_mm_64 0.1380 ms 83.9%
  triton_mm_66 0.1444 ms 80.3%
  triton_mm_62 0.1448 ms 80.0%
  addmm 0.1454 ms 79.7%
  triton_mm_65 0.1562 ms 74.2%
  triton_mm_67 0.1656 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 2.5473 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(3432x7680, 3432x1920, 1920x7680)
  triton_mm_89 0.4324 ms 100.0%
  triton_mm_88 0.4403 ms 98.2%
  triton_mm_90 0.4777 ms 90.5%
  bias_addmm 0.5172 ms 83.6%
  triton_mm_83 0.5337 ms 81.0%
  triton_mm_85 0.5544 ms 78.0%
  triton_mm_81 0.5667 ms 76.3%
  triton_mm_84 0.5876 ms 73.6%
  triton_mm_86 0.6250 ms 69.2%
  triton_mm_82 0.6414 ms 67.4%
SingleProcess AUTOTUNE benchmarking takes 2.7005 seconds and 1.7882 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_108 0.4300 ms 100.0%
  addmm 0.4300 ms 100.0%
  triton_mm_107 0.4508 ms 95.4%
  triton_mm_109 0.4526 ms 95.0%
  bias_addmm 0.4943 ms 87.0%
  triton_mm_102 0.5343 ms 80.5%
  triton_mm_100 0.5560 ms 77.3%
  triton_mm_104 0.5587 ms 77.0%
  triton_mm_103 0.5597 ms 76.8%
  triton_mm_105 0.6246 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7051 seconds and 0.2116 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_122 0.0473 ms 100.0%
  triton_mm_118 0.0485 ms 97.5%
  triton_mm_126 0.0486 ms 97.2%
  triton_mm_117 0.0493 ms 95.8%
  triton_mm_121 0.0499 ms 94.8%
  triton_mm_111 0.0503 ms 94.1%
  bias_addmm 0.0507 ms 93.3%
  triton_mm_112 0.0507 ms 93.3%
  triton_mm_114 0.0513 ms 92.3%
  triton_mm_113 0.0539 ms 87.8%
SingleProcess AUTOTUNE benchmarking takes 2.2187 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x5760, 3432x1920, 1920x5760)
  triton_mm_144 0.3369 ms 100.0%
  triton_mm_143 0.3464 ms 97.2%
  triton_mm_145 0.3709 ms 90.8%
  triton_mm_138 0.3986 ms 84.5%
  bias_addmm 0.4000 ms 84.2%
  triton_mm_140 0.4180 ms 80.6%
  triton_mm_136 0.4260 ms 79.1%
  triton_mm_139 0.4458 ms 75.6%
  addmm 0.4596 ms 73.3%
  triton_mm_141 0.4756 ms 70.8%
SingleProcess AUTOTUNE benchmarking takes 2.6635 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_163 0.1160 ms 100.0%
  triton_mm_162 0.1189 ms 97.6%
  triton_mm_164 0.1253 ms 92.6%
  bias_addmm 0.1282 ms 90.5%
  triton_mm_157 0.1380 ms 84.1%
  triton_mm_159 0.1445 ms 80.3%
  triton_mm_155 0.1449 ms 80.1%
  addmm 0.1452 ms 79.9%
  triton_mm_158 0.1567 ms 74.0%
  triton_mm_160 0.1651 ms 70.3%
SingleProcess AUTOTUNE benchmarking takes 2.5564 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_201 0.4297 ms 100.0%
  triton_mm_200 0.4505 ms 95.4%
  triton_mm_202 0.4548 ms 94.5%
  addmm 0.4916 ms 87.4%
  bias_addmm 0.4942 ms 86.9%
  triton_mm_195 0.5348 ms 80.3%
  triton_mm_193 0.5587 ms 76.9%
  triton_mm_197 0.5594 ms 76.8%
  triton_mm_196 0.5608 ms 76.6%
  triton_mm_198 0.6250 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7136 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_215 0.0471 ms 100.0%
  triton_mm_219 0.0478 ms 98.5%
  triton_mm_211 0.0483 ms 97.5%
  triton_mm_214 0.0488 ms 96.6%
  triton_mm_210 0.0491 ms 95.9%
  triton_mm_204 0.0501 ms 93.9%
  triton_mm_207 0.0504 ms 93.4%
  triton_mm_205 0.0511 ms 92.1%
  bias_addmm 0.0516 ms 91.2%
  triton_mm_206 0.0533 ms 88.4%
SingleProcess AUTOTUNE benchmarking takes 2.2158 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x1920, 1920x1920)
  triton_mm_256 0.1159 ms 100.0%
  triton_mm_255 0.1181 ms 98.1%
  triton_mm_257 0.1253 ms 92.5%
  triton_mm_250 0.1372 ms 84.4%
  triton_mm_252 0.1437 ms 80.6%
  triton_mm_248 0.1447 ms 80.1%
  addmm 0.1472 ms 78.7%
  bias_addmm 0.1514 ms 76.5%
  triton_mm_251 0.1567 ms 74.0%
  triton_mm_253 0.1649 ms 70.3%
SingleProcess AUTOTUNE benchmarking takes 2.5663 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x1920, 3432x7680, 7680x1920)
  triton_mm_294 0.4315 ms 100.0%
  addmm 0.4316 ms 100.0%
  triton_mm_293 0.4509 ms 95.7%
  triton_mm_295 0.4551 ms 94.8%
  bias_addmm 0.4961 ms 87.0%
  triton_mm_288 0.5364 ms 80.4%
  triton_mm_286 0.5595 ms 77.1%
  triton_mm_290 0.5597 ms 77.1%
  triton_mm_289 0.5613 ms 76.9%
  triton_mm_291 0.6257 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7084 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x11520, 12x1920, 1920x11520)
  triton_mm_308 0.0471 ms 100.0%
  triton_mm_312 0.0479 ms 98.3%
  triton_mm_304 0.0482 ms 97.7%
  triton_mm_307 0.0498 ms 94.7%
  triton_mm_303 0.0498 ms 94.6%
  triton_mm_297 0.0502 ms 93.8%
  triton_mm_298 0.0506 ms 93.1%
  triton_mm_300 0.0512 ms 92.1%
  bias_addmm 0.0517 ms 91.1%
  triton_mm_299 0.0533 ms 88.3%
SingleProcess AUTOTUNE benchmarking takes 2.2194 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x1920, 1920x2048)
  triton_mm_592 0.1175 ms 100.0%
  triton_mm_591 0.1196 ms 98.2%
  triton_mm_593 0.1268 ms 92.6%
  triton_mm_586 0.1426 ms 82.4%
  triton_mm_584 0.1482 ms 79.3%
  triton_mm_588 0.1495 ms 78.6%
  bias_addmm 0.1547 ms 76.0%
  triton_mm_587 0.1644 ms 71.5%
  triton_mm_589 0.1671 ms 70.3%
  triton_mm_585 0.1690 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.5534 seconds and 0.0038 seconds precompiling
AUTOTUNE addmm(3432x2048, 3432x2048, 2048x2048)
  triton_mm_611 0.1231 ms 100.0%
  triton_mm_610 0.1239 ms 99.4%
  triton_mm_612 0.1337 ms 92.1%
  bias_addmm 0.1339 ms 92.0%
  triton_mm_607 0.1536 ms 80.2%
  addmm 0.1536 ms 80.2%
  triton_mm_605 0.1545 ms 79.7%
  triton_mm_603 0.1547 ms 79.6%
  triton_mm_606 0.1721 ms 71.6%
  triton_mm_608 0.1743 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.5806 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x768)
  mm 0.0589 ms 100.0%
  triton_mm_624 0.0603 ms 97.7%
  triton_mm_626 0.0609 ms 96.7%
  triton_mm_622 0.0626 ms 94.1%
  triton_mm_630 0.0630 ms 93.5%
  triton_mm_629 0.0636 ms 92.6%
  triton_mm_625 0.0694 ms 84.9%
  triton_mm_627 0.0721 ms 81.8%
  triton_mm_623 0.0733 ms 80.4%
  triton_mm_620 0.0793 ms 74.3%
SingleProcess AUTOTUNE benchmarking takes 10.2953 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0225 ms 100.0%
  triton_mm_2880 0.0225 ms 99.9%
  triton_mm_2872 0.0228 ms 98.7%
  triton_mm_2868 0.0236 ms 95.4%
  bias_addmm 0.0255 ms 88.0%
  triton_mm_2871 0.0272 ms 82.5%
  triton_mm_2867 0.0273 ms 82.2%
  triton_mm_2865 0.0277 ms 81.0%
  addmm 0.0283 ms 79.5%
  triton_mm_2875 0.0283 ms 79.5%
SingleProcess AUTOTUNE benchmarking takes 7.7406 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2898 1.8483 ms 100.0%
  triton_mm_2897 1.8681 ms 98.9%
  bias_addmm 1.9850 ms 93.1%
  triton_mm_2899 2.0330 ms 90.9%
  triton_mm_2892 2.2671 ms 81.5%
  triton_mm_2894 2.3348 ms 79.2%
  triton_mm_2890 2.4081 ms 76.8%
  addmm 2.4526 ms 75.4%
  triton_mm_2893 2.4644 ms 75.0%
  triton_mm_2895 2.6548 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 11.8910 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x3432, 3432x1920)
  mm 1.6927 ms 100.0%
  triton_mm_2935 1.7326 ms 97.7%
  triton_mm_2936 1.7879 ms 94.7%
  triton_mm_2937 1.9213 ms 88.1%
  triton_mm_2928 2.2230 ms 76.1%
  triton_mm_2930 2.3032 ms 73.5%
  triton_mm_2932 2.3318 ms 72.6%
  triton_mm_2931 2.3676 ms 71.5%
  triton_mm_2933 2.4402 ms 69.4%
  triton_mm_2929 2.4720 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 9.9277 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x32768, 32768x1920)
  mm 1.6864 ms 100.0%
  triton_mm_2916 1.8597 ms 90.7%
  triton_mm_2917 1.8952 ms 89.0%
  triton_mm_2918 1.8989 ms 88.8%
  triton_mm_2912 2.3367 ms 72.2%
  triton_mm_2911 2.4562 ms 68.7%
  triton_mm_2909 2.4929 ms 67.6%
  triton_mm_2910 2.5526 ms 66.1%
  triton_mm_2913 2.5562 ms 66.0%
  triton_mm_2914 2.6429 ms 63.8%
SingleProcess AUTOTUNE benchmarking takes 9.9120 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3006 0.4228 ms 100.0%
  triton_mm_3007 0.4364 ms 96.9%
  triton_mm_3008 0.4609 ms 91.7%
  mm 0.4787 ms 88.3%
  triton_mm_3003 0.5265 ms 80.3%
  triton_mm_2999 0.5377 ms 78.6%
  triton_mm_3001 0.5412 ms 78.1%
  triton_mm_3004 0.5679 ms 74.4%
  triton_mm_3002 0.5754 ms 73.5%
  triton_mm_3000 0.5844 ms 72.3%
SingleProcess AUTOTUNE benchmarking takes 9.9506 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x12, 12x1920)
  triton_mm_2962 0.0160 ms 100.0%
  triton_mm_2961 0.0164 ms 97.7%
  triton_mm_2963 0.0166 ms 96.7%
  triton_mm_2960 0.0171 ms 93.8%
  triton_mm_2965 0.0172 ms 93.1%
  triton_mm_2966 0.0173 ms 92.8%
  triton_mm_2964 0.0174 ms 91.9%
  triton_mm_2967 0.0174 ms 91.9%
  triton_mm_2959 0.0181 ms 88.5%
  triton_mm_2969 0.0184 ms 87.0%
SingleProcess AUTOTUNE benchmarking takes 5.3799 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0218 ms 100.0%
  triton_mm_2942 0.0238 ms 91.4%
  triton_mm_2950 0.0246 ms 88.6%
  mm 0.0278 ms 78.5%
  triton_mm_2954 0.0284 ms 76.8%
  triton_mm_2940 0.0342 ms 63.8%
  triton_mm_2941 0.0345 ms 63.2%
  triton_mm_2945 0.0372 ms 58.5%
  triton_mm_2949 0.0396 ms 55.0%
  triton_mm_2939 0.0414 ms 52.6%
SingleProcess AUTOTUNE benchmarking takes 2.0850 seconds and 0.0008 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_2987 0.4186 ms 100.0%
  triton_mm_2988 0.4220 ms 99.2%
  mm 0.4441 ms 94.3%
  triton_mm_2989 0.4589 ms 91.2%
  triton_mm_2982 0.5239 ms 79.9%
  triton_mm_2984 0.5334 ms 78.5%
  triton_mm_2980 0.5337 ms 78.4%
  triton_mm_2983 0.5691 ms 73.6%
  triton_mm_2981 0.5918 ms 70.7%
  triton_mm_2985 0.6059 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.2278 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x3432, 3432x1920)
  mm 0.4174 ms 100.0%
  triton_mm_3044 0.4214 ms 99.1%
  triton_mm_3045 0.4391 ms 95.1%
  triton_mm_3046 0.4658 ms 89.6%
  triton_mm_3037 0.5219 ms 80.0%
  triton_mm_3041 0.5294 ms 78.9%
  triton_mm_3039 0.5310 ms 78.6%
  triton_mm_3040 0.5595 ms 74.6%
  triton_mm_3042 0.5707 ms 73.1%
  triton_mm_3038 0.5752 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 12.9823 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4020 ms 100.0%
  triton_mm_3025 0.4299 ms 93.5%
  triton_mm_3026 0.4422 ms 90.9%
  triton_mm_3027 0.4473 ms 89.9%
  triton_mm_3020 0.5240 ms 76.7%
  triton_mm_3018 0.5354 ms 75.1%
  triton_mm_3022 0.5431 ms 74.0%
  triton_mm_3021 0.5481 ms 73.3%
  triton_mm_3019 0.5975 ms 67.3%
  triton_mm_3023 0.6202 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 9.9825 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3217 ms 100.0%
  triton_mm_3120 0.3259 ms 98.7%
  triton_mm_3121 0.3390 ms 94.9%
  triton_mm_3113 0.3992 ms 80.6%
  triton_mm_3115 0.4025 ms 79.9%
  triton_mm_3117 0.4042 ms 79.6%
  triton_mm_3122 0.4112 ms 78.2%
  triton_mm_3116 0.4268 ms 75.4%
  triton_mm_3118 0.4362 ms 73.8%
  triton_mm_3114 0.4393 ms 73.2%
SingleProcess AUTOTUNE benchmarking takes 9.9562 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3101 0.3224 ms 100.0%
  triton_mm_3102 0.3246 ms 99.3%
  mm 0.3291 ms 98.0%
  triton_mm_3103 0.3380 ms 95.4%
  triton_mm_3096 0.3942 ms 81.8%
  triton_mm_3094 0.4037 ms 79.8%
  triton_mm_3098 0.4097 ms 78.7%
  triton_mm_3097 0.4111 ms 78.4%
  triton_mm_3095 0.4488 ms 71.8%
  triton_mm_3099 0.4653 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 9.8125 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3082 0.1402 ms 100.0%
  triton_mm_3075 0.1461 ms 96.0%
  triton_mm_3077 0.1470 ms 95.4%
  triton_mm_3079 0.1476 ms 95.0%
  triton_mm_3083 0.1477 ms 94.9%
  triton_mm_3078 0.1647 ms 85.1%
  triton_mm_3080 0.1675 ms 83.7%
  triton_mm_3076 0.1689 ms 83.0%
  mm 0.1767 ms 79.3%
  triton_mm_3084 0.1805 ms 77.6%
SingleProcess AUTOTUNE benchmarking takes 9.9481 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1103 ms 100.0%
  triton_mm_3063 0.1119 ms 98.6%
  triton_mm_3064 0.1133 ms 97.3%
  triton_mm_3065 0.1205 ms 91.5%
  triton_mm_3058 0.1343 ms 82.1%
  triton_mm_3056 0.1379 ms 80.0%
  triton_mm_3060 0.1393 ms 79.2%
  triton_mm_3059 0.1483 ms 74.4%
  triton_mm_3057 0.1545 ms 71.4%
  triton_mm_3061 0.1607 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 14.5195 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x12, 12x1920)
  triton_mm_3152 0.0322 ms 100.0%
  triton_mm_3149 0.0326 ms 98.7%
  triton_mm_3151 0.0328 ms 98.0%
  triton_mm_3148 0.0333 ms 96.6%
  triton_mm_3145 0.0336 ms 95.8%
  triton_mm_3147 0.0336 ms 95.7%
  triton_mm_3154 0.0337 ms 95.4%
  triton_mm_3150 0.0343 ms 93.7%
  triton_mm_3146 0.0348 ms 92.3%
  triton_mm_3155 0.0352 ms 91.4%
SingleProcess AUTOTUNE benchmarking takes 5.2893 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3191 0.4234 ms 100.0%
  mm 0.4305 ms 98.3%
  triton_mm_3192 0.4377 ms 96.7%
  triton_mm_3193 0.4631 ms 91.4%
  triton_mm_3188 0.5277 ms 80.2%
  triton_mm_3184 0.5386 ms 78.6%
  triton_mm_3186 0.5422 ms 78.1%
  triton_mm_3189 0.5723 ms 74.0%
  triton_mm_3187 0.5775 ms 73.3%
  triton_mm_3185 0.5849 ms 72.4%
SingleProcess AUTOTUNE benchmarking takes 2.5658 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3131 0.0528 ms 100.0%
  mm 0.0530 ms 99.7%
  triton_mm_3127 0.0536 ms 98.5%
  triton_mm_3135 0.0584 ms 90.4%
  triton_mm_3139 0.0680 ms 77.7%
  triton_mm_3126 0.0842 ms 62.7%
  triton_mm_3125 0.0883 ms 59.8%
  triton_mm_3130 0.0946 ms 55.9%
  triton_mm_3134 0.1009 ms 52.4%
  triton_mm_3124 0.1032 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 6.5016 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3172 0.4165 ms 100.0%
  triton_mm_3173 0.4231 ms 98.4%
  mm 0.4430 ms 94.0%
  triton_mm_3174 0.4609 ms 90.4%
  triton_mm_3167 0.5240 ms 79.5%
  triton_mm_3165 0.5339 ms 78.0%
  triton_mm_3169 0.5349 ms 77.9%
  triton_mm_3168 0.5713 ms 72.9%
  triton_mm_3166 0.5918 ms 70.4%
  triton_mm_3170 0.6135 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 2.5693 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.4012 ms 100.0%
  triton_mm_3210 0.4293 ms 93.5%
  triton_mm_3211 0.4426 ms 90.7%
  triton_mm_3212 0.4495 ms 89.3%
  triton_mm_3205 0.5255 ms 76.4%
  triton_mm_3203 0.5351 ms 75.0%
  triton_mm_3207 0.5435 ms 73.8%
  triton_mm_3206 0.5504 ms 72.9%
  triton_mm_3204 0.5971 ms 67.2%
  triton_mm_3208 0.6206 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.5802 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3224 ms 100.0%
  triton_mm_3305 0.3249 ms 99.2%
  triton_mm_3306 0.3398 ms 94.9%
  triton_mm_3307 0.3587 ms 89.9%
  triton_mm_3298 0.3992 ms 80.8%
  triton_mm_3300 0.4026 ms 80.1%
  triton_mm_3302 0.4035 ms 79.9%
  triton_mm_3301 0.4288 ms 75.2%
  triton_mm_3303 0.4360 ms 73.9%
  triton_mm_3299 0.4397 ms 73.3%
SingleProcess AUTOTUNE benchmarking takes 2.5412 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3286 0.3221 ms 100.0%
  triton_mm_3287 0.3245 ms 99.2%
  mm 0.3299 ms 97.6%
  triton_mm_3288 0.3390 ms 95.0%
  triton_mm_3281 0.3949 ms 81.6%
  triton_mm_3279 0.4034 ms 79.8%
  triton_mm_3283 0.4111 ms 78.3%
  triton_mm_3282 0.4121 ms 78.2%
  triton_mm_3280 0.4498 ms 71.6%
  triton_mm_3284 0.4660 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.5559 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3267 0.1397 ms 100.0%
  triton_mm_3260 0.1458 ms 95.9%
  triton_mm_3264 0.1479 ms 94.5%
  triton_mm_3268 0.1481 ms 94.4%
  triton_mm_3262 0.1481 ms 94.4%
  triton_mm_3269 0.1499 ms 93.2%
  triton_mm_3263 0.1640 ms 85.2%
  triton_mm_3265 0.1675 ms 83.5%
  triton_mm_3261 0.1694 ms 82.5%
  mm 0.1792 ms 78.0%
SingleProcess AUTOTUNE benchmarking takes 2.4492 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1107 ms 100.0%
  triton_mm_3248 0.1123 ms 98.6%
  triton_mm_3249 0.1136 ms 97.4%
  triton_mm_3250 0.1208 ms 91.6%
  triton_mm_3243 0.1348 ms 82.1%
  triton_mm_3241 0.1391 ms 79.6%
  triton_mm_3245 0.1399 ms 79.1%
  triton_mm_3244 0.1483 ms 74.6%
  triton_mm_3242 0.1549 ms 71.4%
  triton_mm_3246 0.1611 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.4473 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x7680)
  triton_mm_3376 0.4229 ms 100.0%
  mm 0.4300 ms 98.4%
  triton_mm_3377 0.4370 ms 96.8%
  triton_mm_3378 0.4631 ms 91.3%
  triton_mm_3373 0.5278 ms 80.1%
  triton_mm_3369 0.5417 ms 78.1%
  triton_mm_3371 0.5435 ms 77.8%
  triton_mm_3374 0.5723 ms 73.9%
  triton_mm_3372 0.5778 ms 73.2%
  triton_mm_3370 0.5833 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.5773 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3312 0.0527 ms 100.0%
  mm 0.0530 ms 99.4%
  triton_mm_3316 0.0531 ms 99.2%
  triton_mm_3320 0.0574 ms 91.7%
  triton_mm_3324 0.0680 ms 77.5%
  triton_mm_3311 0.0843 ms 62.5%
  triton_mm_3310 0.0876 ms 60.1%
  triton_mm_3315 0.0945 ms 55.7%
  triton_mm_3319 0.1015 ms 51.9%
  triton_mm_3309 0.1032 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 2.1548 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3357 0.4185 ms 100.0%
  triton_mm_3358 0.4254 ms 98.4%
  mm 0.4396 ms 95.2%
  triton_mm_3359 0.4623 ms 90.5%
  triton_mm_3352 0.5251 ms 79.7%
  triton_mm_3350 0.5344 ms 78.3%
  triton_mm_3354 0.5353 ms 78.2%
  triton_mm_3353 0.5717 ms 73.2%
  triton_mm_3351 0.5932 ms 70.5%
  triton_mm_3355 0.6133 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 2.5743 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x7680, 7680x1920)
  mm 0.3980 ms 100.0%
  triton_mm_3395 0.4308 ms 92.4%
  triton_mm_3396 0.4438 ms 89.7%
  triton_mm_3397 0.4503 ms 88.4%
  triton_mm_3390 0.5272 ms 75.5%
  triton_mm_3388 0.5325 ms 74.7%
  triton_mm_3392 0.5448 ms 73.1%
  triton_mm_3391 0.5520 ms 72.1%
  triton_mm_3389 0.5994 ms 66.4%
  triton_mm_3393 0.6225 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 2.5758 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x3432, 3432x1920)
  mm 0.3242 ms 100.0%
  triton_mm_3490 0.3254 ms 99.6%
  triton_mm_3491 0.3410 ms 95.1%
  triton_mm_3492 0.3594 ms 90.2%
  triton_mm_3483 0.3998 ms 81.1%
  triton_mm_3485 0.4048 ms 80.1%
  triton_mm_3487 0.4060 ms 79.9%
  triton_mm_3486 0.4297 ms 75.4%
  triton_mm_3488 0.4365 ms 74.3%
  triton_mm_3484 0.4396 ms 73.7%
SingleProcess AUTOTUNE benchmarking takes 2.5449 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x5760, 5760x1920)
  triton_mm_3471 0.3228 ms 100.0%
  triton_mm_3472 0.3248 ms 99.4%
  mm 0.3304 ms 97.7%
  triton_mm_3473 0.3404 ms 94.8%
  triton_mm_3466 0.3948 ms 81.8%
  triton_mm_3464 0.4039 ms 79.9%
  triton_mm_3468 0.4118 ms 78.4%
  triton_mm_3467 0.4124 ms 78.3%
  triton_mm_3465 0.4501 ms 71.7%
  triton_mm_3469 0.4678 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.5530 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3432, 3432x1920)
  triton_mm_3452 0.1406 ms 100.0%
  triton_mm_3453 0.1472 ms 95.5%
  triton_mm_3445 0.1473 ms 95.4%
  triton_mm_3447 0.1475 ms 95.3%
  triton_mm_3449 0.1482 ms 94.8%
  triton_mm_3454 0.1493 ms 94.1%
  triton_mm_3448 0.1647 ms 85.3%
  triton_mm_3450 0.1681 ms 83.6%
  triton_mm_3446 0.1703 ms 82.6%
  mm 0.1796 ms 78.3%
SingleProcess AUTOTUNE benchmarking takes 2.4500 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x1920)
  mm 0.1099 ms 100.0%
  triton_mm_3433 0.1125 ms 97.7%
  triton_mm_3434 0.1144 ms 96.0%
  triton_mm_3435 0.1218 ms 90.2%
  triton_mm_3428 0.1355 ms 81.1%
  triton_mm_3426 0.1393 ms 78.9%
  triton_mm_3430 0.1407 ms 78.1%
  triton_mm_3429 0.1491 ms 73.7%
  triton_mm_3427 0.1557 ms 70.6%
  triton_mm_3431 0.1614 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.4413 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  triton_mm_3497 0.0525 ms 100.0%
  mm 0.0528 ms 99.6%
  triton_mm_3501 0.0530 ms 99.1%
  triton_mm_3505 0.0575 ms 91.4%
  triton_mm_3509 0.0680 ms 77.2%
  triton_mm_3496 0.0844 ms 62.3%
  triton_mm_3495 0.0878 ms 59.9%
  triton_mm_3500 0.0946 ms 55.6%
  triton_mm_3504 0.1016 ms 51.7%
  triton_mm_3494 0.1033 ms 50.9%
SingleProcess AUTOTUNE benchmarking takes 2.1518 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3542 0.4194 ms 100.0%
  triton_mm_3543 0.4248 ms 98.7%
  mm 0.4439 ms 94.5%
  triton_mm_3544 0.4629 ms 90.6%
  triton_mm_3537 0.5277 ms 79.5%
  triton_mm_3535 0.5355 ms 78.3%
  triton_mm_3539 0.5392 ms 77.8%
  triton_mm_3538 0.5753 ms 72.9%
  triton_mm_3536 0.5940 ms 70.6%
  triton_mm_3540 0.6078 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.5685 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0528 ms 100.0%
  triton_mm_3686 0.0529 ms 99.9%
  triton_mm_3682 0.0534 ms 98.9%
  triton_mm_3690 0.0575 ms 91.9%
  triton_mm_3694 0.0681 ms 77.6%
  triton_mm_3681 0.0850 ms 62.1%
  triton_mm_3680 0.0884 ms 59.8%
  triton_mm_3685 0.0946 ms 55.8%
  triton_mm_3689 0.1017 ms 52.0%
  triton_mm_3679 0.1040 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 2.1508 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3727 0.4168 ms 100.0%
  triton_mm_3728 0.4240 ms 98.3%
  mm 0.4427 ms 94.2%
  triton_mm_3729 0.4621 ms 90.2%
  triton_mm_3722 0.5251 ms 79.4%
  triton_mm_3720 0.5348 ms 77.9%
  triton_mm_3724 0.5350 ms 77.9%
  triton_mm_3723 0.5751 ms 72.5%
  triton_mm_3721 0.5925 ms 70.3%
  triton_mm_3725 0.6134 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 2.5717 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(12x11520, 11520x1920)
  mm 0.0527 ms 100.0%
  triton_mm_3871 0.0528 ms 99.9%
  triton_mm_3867 0.0532 ms 99.0%
  triton_mm_3875 0.0585 ms 90.1%
  triton_mm_3879 0.0691 ms 76.3%
  triton_mm_3866 0.0843 ms 62.6%
  triton_mm_3865 0.0876 ms 60.1%
  triton_mm_3870 0.0946 ms 55.7%
  triton_mm_3874 0.1016 ms 51.9%
  triton_mm_3864 0.1033 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 2.1593 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x1920, 1920x7680)
  triton_mm_3912 0.4191 ms 100.0%
  triton_mm_3913 0.4239 ms 98.9%
  mm 0.4418 ms 94.9%
  triton_mm_3914 0.4620 ms 90.7%
  triton_mm_3907 0.5276 ms 79.4%
  triton_mm_3909 0.5358 ms 78.2%
  triton_mm_3905 0.5383 ms 77.9%
  triton_mm_3908 0.5736 ms 73.1%
  triton_mm_3906 0.5934 ms 70.6%
  triton_mm_3910 0.6129 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.5757 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x768, 768x2048)
  mm 0.0545 ms 100.0%
  triton_mm_7427 0.0549 ms 99.3%
  triton_mm_7428 0.0572 ms 95.4%
  triton_mm_7424 0.0634 ms 86.0%
  triton_mm_7420 0.0639 ms 85.3%
  triton_mm_7421 0.0682 ms 80.0%
  triton_mm_7429 0.0688 ms 79.3%
  triton_mm_7425 0.0715 ms 76.3%
  triton_mm_7422 0.0800 ms 68.1%
  triton_mm_7423 0.0842 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 10.4380 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x2048)
  mm 0.1166 ms 100.0%
  triton_mm_7465 0.1198 ms 97.4%
  triton_mm_7466 0.1244 ms 93.7%
  triton_mm_7467 0.1307 ms 89.3%
  triton_mm_7460 0.1516 ms 76.9%
  triton_mm_7458 0.1518 ms 76.8%
  triton_mm_7462 0.1562 ms 74.7%
  triton_mm_7459 0.1637 ms 71.2%
  triton_mm_7461 0.1637 ms 71.2%
  triton_mm_7463 0.1669 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 10.2473 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x2048)
  mm 0.1278 ms 100.0%
  triton_mm_7484 0.1440 ms 88.7%
  triton_mm_7485 0.1487 ms 86.0%
  triton_mm_7486 0.1539 ms 83.1%
  triton_mm_7481 0.1599 ms 80.0%
  triton_mm_7477 0.1620 ms 78.9%
  triton_mm_7479 0.1661 ms 77.0%
  triton_mm_7480 0.1674 ms 76.4%
  triton_mm_7482 0.1700 ms 75.2%
  triton_mm_7478 0.1713 ms 74.6%
SingleProcess AUTOTUNE benchmarking takes 10.0939 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x3432, 3432x1920)
  mm 0.1300 ms 100.0%
  triton_mm_7522 0.1439 ms 90.3%
  triton_mm_7523 0.1495 ms 86.9%
  triton_mm_7519 0.1501 ms 86.6%
  triton_mm_7515 0.1508 ms 86.2%
  triton_mm_7524 0.1512 ms 85.9%
  triton_mm_7517 0.1545 ms 84.1%
  triton_mm_7518 0.1649 ms 78.8%
  triton_mm_7520 0.1703 ms 76.3%
  triton_mm_7516 0.1709 ms 76.0%
SingleProcess AUTOTUNE benchmarking takes 10.1774 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x3432, 3432x2048)
  mm 0.0563 ms 100.0%
  triton_mm_7448 0.0571 ms 98.5%
  triton_mm_7447 0.0609 ms 92.5%
  triton_mm_7441 0.0657 ms 85.7%
  triton_mm_7439 0.0699 ms 80.6%
  triton_mm_7442 0.0705 ms 79.8%
  triton_mm_7446 0.0710 ms 79.3%
  triton_mm_7443 0.0725 ms 77.7%
  triton_mm_7440 0.0732 ms 76.9%
  triton_mm_7444 0.0741 ms 75.9%
SingleProcess AUTOTUNE benchmarking takes 10.1055 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3432x2048, 2048x1920)
  triton_mm_7503 0.1182 ms 100.0%
  mm 0.1184 ms 99.8%
  triton_mm_7504 0.1253 ms 94.3%
  triton_mm_7505 0.1294 ms 91.3%
  triton_mm_7500 0.1476 ms 80.1%
  triton_mm_7496 0.1479 ms 80.0%
  triton_mm_7498 0.1499 ms 78.9%
  triton_mm_7499 0.1582 ms 74.7%
  triton_mm_7497 0.1639 ms 72.1%
  triton_mm_7501 0.1687 ms 70.1%
SingleProcess AUTOTUNE benchmarking takes 12.9942 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x3420, 3420x28)
  mm 0.0306 ms 100.0%
  triton_mm_8637 0.0350 ms 87.4%
  triton_mm_8638 0.0392 ms 77.9%
  triton_mm_8636 0.0409 ms 74.7%
  triton_mm_8645 0.0411 ms 74.4%
  triton_mm_8642 0.0422 ms 72.4%
  triton_mm_8646 0.0438 ms 69.8%
  triton_mm_8639 0.0445 ms 68.7%
  triton_mm_8650 0.0472 ms 64.8%
  triton_mm_8644 0.0475 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 7.3379 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2562 ms 100.0%
  mm 1.4447 ms 87.0%
  triton_mm_8754 1.5240 ms 82.4%
  triton_mm_8758 1.5462 ms 81.2%
  triton_mm_8750 1.5554 ms 80.8%
  triton_mm_8755 1.6624 ms 75.6%
  triton_mm_8751 1.6668 ms 75.4%
  triton_mm_8752 1.6883 ms 74.4%
  triton_mm_8747 2.0726 ms 60.6%
  triton_mm_8748 2.1284 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2216 seconds and 0.0017 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2523 ms 100.0%
  mm 1.4472 ms 86.5%
  triton_mm_8847 1.5362 ms 81.5%
  triton_mm_8851 1.5368 ms 81.5%
  triton_mm_8843 1.5750 ms 79.5%
  triton_mm_8848 1.6604 ms 75.4%
  triton_mm_8845 1.6807 ms 74.5%
  triton_mm_8844 1.6812 ms 74.5%
  triton_mm_8840 2.0656 ms 60.6%
  triton_mm_8841 2.1153 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2109 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8943 1.2549 ms 100.0%
  mm 1.2841 ms 97.7%
  triton_mm_8940 1.5401 ms 81.5%
  triton_mm_8936 1.5421 ms 81.4%
  triton_mm_8944 1.5575 ms 80.6%
  triton_mm_8941 1.6654 ms 75.4%
  triton_mm_8937 1.6850 ms 74.5%
  triton_mm_8938 1.6879 ms 74.3%
  triton_mm_8933 2.0644 ms 60.8%
  triton_mm_8934 2.1436 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2168 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10152 1.2526 ms 100.0%
  mm 1.2832 ms 97.6%
  triton_mm_10149 1.5353 ms 81.6%
  triton_mm_10153 1.5598 ms 80.3%
  triton_mm_10145 1.5719 ms 79.7%
  triton_mm_10150 1.6599 ms 75.5%
  triton_mm_10147 1.6818 ms 74.5%
  triton_mm_10146 1.6996 ms 73.7%
  triton_mm_10142 2.0578 ms 60.9%
  triton_mm_10143 2.1387 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2183 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10896 1.2551 ms 100.0%
  mm 1.2870 ms 97.5%
  triton_mm_10893 1.5399 ms 81.5%
  triton_mm_10889 1.5573 ms 80.6%
  triton_mm_10897 1.5582 ms 80.5%
  triton_mm_10890 1.6656 ms 75.4%
  triton_mm_10894 1.6843 ms 74.5%
  triton_mm_10891 1.7020 ms 73.7%
  triton_mm_10886 2.0644 ms 60.8%
  triton_mm_10887 2.1236 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2244 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11082 1.2658 ms 100.0%
  mm 1.2867 ms 98.4%
  triton_mm_11079 1.5361 ms 82.4%
  triton_mm_11075 1.5549 ms 81.4%
  triton_mm_11083 1.5669 ms 80.8%
  triton_mm_11080 1.6608 ms 76.2%
  triton_mm_11077 1.6972 ms 74.6%
  triton_mm_11076 1.6993 ms 74.5%
  triton_mm_11072 2.0793 ms 60.9%
  triton_mm_11073 2.1374 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2278 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3709 ms 100.0%
  mm 0.3797 ms 97.7%
  triton_mm_11504 0.4117 ms 90.1%
  triton_mm_11508 0.4207 ms 88.2%
  triton_mm_11505 0.4439 ms 83.5%
  triton_mm_11506 0.4535 ms 81.8%
  triton_mm_11512 0.4575 ms 81.1%
  triton_mm_11509 0.4644 ms 79.9%
  triton_mm_11501 0.5584 ms 66.4%
  triton_mm_11502 0.5749 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.0578 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3933 ms 100.0%
  mm 0.4020 ms 97.9%
  triton_mm_11523 0.4331 ms 90.8%
  triton_mm_11527 0.4551 ms 86.4%
  triton_mm_11524 0.4670 ms 84.2%
  triton_mm_11525 0.4767 ms 82.5%
  triton_mm_11531 0.4869 ms 80.8%
  triton_mm_11528 0.4900 ms 80.3%
  triton_mm_11520 0.5865 ms 67.1%
  triton_mm_11521 0.6108 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.0660 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0740 ms 100.0%
  triton_mm_8668 0.0751 ms 98.6%
  triton_mm_8659 0.0761 ms 97.2%
  triton_mm_8662 0.0763 ms 97.1%
  triton_mm_8654 0.0772 ms 95.9%
  triton_mm_8666 0.0777 ms 95.2%
  mm 0.0778 ms 95.2%
  triton_mm_8665 0.0780 ms 94.9%
  triton_mm_8661 0.0781 ms 94.8%
  triton_mm_8653 0.0787 ms 94.0%
SingleProcess AUTOTUNE benchmarking takes 1.8912 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0316 ms 100.0%
  triton_mm_8676 0.0316 ms 100.0%
  triton_mm_8677 0.0316 ms 99.8%
  triton_mm_8681 0.0319 ms 99.0%
  triton_mm_8678 0.0328 ms 96.3%
  triton_mm_8680 0.0328 ms 96.2%
  triton_mm_8673 0.0346 ms 91.4%
  triton_mm_8672 0.0348 ms 90.7%
  triton_mm_8684 0.0361 ms 87.6%
  triton_mm_8671 0.0361 ms 87.4%
SingleProcess AUTOTUNE benchmarking takes 2.1158 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9647 ms 100.0%
  triton_mm_8702 0.9679 ms 99.7%
  addmm 1.0930 ms 88.3%
  triton_mm_8695 1.1604 ms 83.1%
  triton_mm_8699 1.1654 ms 82.8%
  triton_mm_8703 1.2178 ms 79.2%
  triton_mm_8700 1.2692 ms 76.0%
  triton_mm_8696 1.2698 ms 76.0%
  triton_mm_8697 1.2872 ms 74.9%
  triton_mm_8692 1.5739 ms 61.3%
SingleProcess AUTOTUNE benchmarking takes 2.3065 seconds and 0.0017 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3268 ms 100.0%
  mm 0.3370 ms 97.0%
  triton_mm_8714 0.3929 ms 83.2%
  triton_mm_8718 0.3985 ms 82.0%
  triton_mm_8722 0.4014 ms 81.4%
  triton_mm_8715 0.4294 ms 76.1%
  triton_mm_8716 0.4331 ms 75.5%
  triton_mm_8719 0.4343 ms 75.2%
  triton_mm_8711 0.5240 ms 62.4%
  triton_mm_8712 0.5398 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0468 seconds and 0.0025 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2589 ms 100.0%
  mm 1.2664 ms 99.4%
  triton_mm_8777 1.4879 ms 84.6%
  triton_mm_8769 1.5993 ms 78.7%
  triton_mm_8773 1.6196 ms 77.7%
  triton_mm_8770 1.6788 ms 75.0%
  triton_mm_8774 1.6792 ms 75.0%
  triton_mm_8771 1.6814 ms 74.9%
  triton_mm_8766 2.1188 ms 59.4%
  triton_mm_8767 2.1366 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.2277 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9293 0.0746 ms 100.0%
  triton_mm_9298 0.0759 ms 98.3%
  triton_mm_9289 0.0761 ms 98.1%
  triton_mm_9292 0.0762 ms 97.9%
  triton_mm_9296 0.0772 ms 96.7%
  triton_mm_9284 0.0777 ms 96.0%
  mm 0.0779 ms 95.8%
  triton_mm_9283 0.0780 ms 95.6%
  triton_mm_9291 0.0781 ms 95.5%
  triton_mm_9295 0.0790 ms 94.4%
SingleProcess AUTOTUNE benchmarking takes 1.8983 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9334 1.2605 ms 100.0%
  mm 1.2661 ms 99.6%
  triton_mm_9335 1.4874 ms 84.7%
  triton_mm_9327 1.5982 ms 78.9%
  triton_mm_9331 1.6201 ms 77.8%
  triton_mm_9329 1.6721 ms 75.4%
  triton_mm_9328 1.6787 ms 75.1%
  triton_mm_9332 1.6931 ms 74.4%
  triton_mm_9324 2.1189 ms 59.5%
  triton_mm_9325 2.1358 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2413 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_9353 0.9712 ms 100.0%
  bias_addmm 0.9747 ms 99.6%
  addmm 1.1035 ms 88.0%
  triton_mm_9350 1.1651 ms 83.4%
  triton_mm_9346 1.1734 ms 82.8%
  triton_mm_9354 1.2283 ms 79.1%
  triton_mm_9351 1.2684 ms 76.6%
  triton_mm_9347 1.2829 ms 75.7%
  triton_mm_9348 1.2890 ms 75.3%
  triton_mm_9343 1.5743 ms 61.7%
SingleProcess AUTOTUNE benchmarking takes 2.3016 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9372 0.3292 ms 100.0%
  mm 0.3391 ms 97.1%
  triton_mm_9365 0.3937 ms 83.6%
  triton_mm_9369 0.3985 ms 82.6%
  triton_mm_9373 0.4022 ms 81.9%
  triton_mm_9366 0.4308 ms 76.4%
  triton_mm_9367 0.4329 ms 76.0%
  triton_mm_9370 0.4348 ms 75.7%
  triton_mm_9362 0.5226 ms 63.0%
  triton_mm_9363 0.5440 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0640 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0391 ms 100.0%
  triton_mm_11428 0.0393 ms 99.4%
  triton_mm_11422 0.0394 ms 99.3%
  triton_mm_11424 0.0395 ms 99.0%
  triton_mm_11437 0.0400 ms 97.8%
  triton_mm_11432 0.0405 ms 96.4%
  triton_mm_11425 0.0414 ms 94.4%
  triton_mm_11435 0.0418 ms 93.6%
  triton_mm_11431 0.0443 ms 88.2%
  mm 0.0449 ms 87.0%
SingleProcess AUTOTUNE benchmarking takes 1.8797 seconds and 0.0012 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3873 ms 100.0%
  bias_addmm 5.4153 ms 99.5%
  addmm 6.0574 ms 88.9%
  triton_mm_11489 6.6140 ms 81.5%
  triton_mm_11493 6.6679 ms 80.8%
  triton_mm_11485 6.6760 ms 80.7%
  triton_mm_11490 7.1776 ms 75.1%
  triton_mm_11486 7.2710 ms 74.1%
  triton_mm_11487 7.2936 ms 73.9%
  triton_mm_11482 8.8277 ms 61.0%
SingleProcess AUTOTUNE benchmarking takes 2.9803 seconds and 0.0014 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1752 ms 100.0%
  bias_addmm 0.1767 ms 99.2%
  addmm 0.1820 ms 96.3%
  triton_mm_11543 0.1924 ms 91.1%
  triton_mm_11547 0.1935 ms 90.5%
  triton_mm_11542 0.1936 ms 90.5%
  triton_mm_11544 0.1946 ms 90.0%
  triton_mm_11546 0.1962 ms 89.3%
  triton_mm_11550 0.2147 ms 81.6%
  triton_mm_11540 0.2358 ms 74.3%
SingleProcess AUTOTUNE benchmarking takes 2.1254 seconds and 0.0011 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:12, 216.58s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2557 ms 100.0%
  mm 1.4460 ms 86.8%
  triton_mm_8754 1.5296 ms 82.1%
  triton_mm_8758 1.5553 ms 80.7%
  triton_mm_8750 1.5569 ms 80.7%
  triton_mm_8755 1.6689 ms 75.2%
  triton_mm_8751 1.6691 ms 75.2%
  triton_mm_8752 1.6869 ms 74.4%
  triton_mm_8747 2.0492 ms 61.3%
  triton_mm_8748 2.1316 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.2148 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2569 ms 100.0%
  mm 1.2852 ms 97.8%
  triton_mm_8847 1.5267 ms 82.3%
  triton_mm_8843 1.5438 ms 81.4%
  triton_mm_8851 1.5584 ms 80.6%
  triton_mm_8848 1.6673 ms 75.4%
  triton_mm_8844 1.6690 ms 75.3%
  triton_mm_8845 1.6878 ms 74.5%
  triton_mm_8840 2.0722 ms 60.7%
  triton_mm_8841 2.1281 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2316 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9036 1.2582 ms 100.0%
  mm 1.2851 ms 97.9%
  triton_mm_9033 1.5303 ms 82.2%
  triton_mm_9029 1.5587 ms 80.7%
  triton_mm_9037 1.5588 ms 80.7%
  triton_mm_9034 1.6674 ms 75.5%
  triton_mm_9030 1.6698 ms 75.4%
  triton_mm_9031 1.6875 ms 74.6%
  triton_mm_9026 2.0698 ms 60.8%
  triton_mm_9027 2.1475 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2129 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9222 1.2580 ms 100.0%
  mm 1.2912 ms 97.4%
  triton_mm_9219 1.5394 ms 81.7%
  triton_mm_9215 1.5562 ms 80.8%
  triton_mm_9223 1.5581 ms 80.7%
  triton_mm_9220 1.6814 ms 74.8%
  triton_mm_9217 1.6920 ms 74.3%
  triton_mm_9216 1.7016 ms 73.9%
  triton_mm_9212 2.0640 ms 60.9%
  triton_mm_9213 2.1416 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2229 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9873 1.2578 ms 100.0%
  mm 1.2721 ms 98.9%
  triton_mm_9870 1.5443 ms 81.4%
  triton_mm_9874 1.5592 ms 80.7%
  triton_mm_9866 1.5593 ms 80.7%
  triton_mm_9871 1.6666 ms 75.5%
  triton_mm_9867 1.6695 ms 75.3%
  triton_mm_9868 1.7056 ms 73.7%
  triton_mm_9863 2.0916 ms 60.1%
  triton_mm_9864 2.1480 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2225 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10617 1.2680 ms 100.0%
  mm 1.2908 ms 98.2%
  triton_mm_10614 1.5384 ms 82.4%
  triton_mm_10618 1.5535 ms 81.6%
  triton_mm_10610 1.5557 ms 81.5%
  triton_mm_10615 1.6798 ms 75.5%
  triton_mm_10611 1.6832 ms 75.3%
  triton_mm_10612 1.6978 ms 74.7%
  triton_mm_10607 2.0862 ms 60.8%
  triton_mm_10608 2.1414 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2296 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10896 1.2668 ms 100.0%
  mm 1.2915 ms 98.1%
  triton_mm_10893 1.5396 ms 82.3%
  triton_mm_10889 1.5461 ms 81.9%
  triton_mm_10897 1.5683 ms 80.8%
  triton_mm_10894 1.6811 ms 75.4%
  triton_mm_10890 1.6825 ms 75.3%
  triton_mm_10891 1.7023 ms 74.4%
  triton_mm_10886 2.0849 ms 60.8%
  triton_mm_10887 2.1415 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2359 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10989 1.2673 ms 100.0%
  mm 1.2929 ms 98.0%
  triton_mm_10986 1.5378 ms 82.4%
  triton_mm_10982 1.5558 ms 81.5%
  triton_mm_10990 1.5703 ms 80.7%
  triton_mm_10987 1.6817 ms 75.4%
  triton_mm_10984 1.6991 ms 74.6%
  triton_mm_10983 1.7023 ms 74.4%
  triton_mm_10979 2.0861 ms 60.7%
  triton_mm_10980 2.1410 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2281 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3742 ms 100.0%
  mm 0.3824 ms 97.9%
  triton_mm_11504 0.4155 ms 90.1%
  triton_mm_11508 0.4223 ms 88.6%
  triton_mm_11505 0.4464 ms 83.8%
  triton_mm_11506 0.4540 ms 82.4%
  triton_mm_11512 0.4588 ms 81.6%
  triton_mm_11509 0.4683 ms 79.9%
  triton_mm_11501 0.5637 ms 66.4%
  triton_mm_11502 0.5803 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.0576 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3962 ms 100.0%
  mm 0.4049 ms 97.8%
  triton_mm_11523 0.4363 ms 90.8%
  triton_mm_11527 0.4596 ms 86.2%
  triton_mm_11524 0.4712 ms 84.1%
  triton_mm_11525 0.4779 ms 82.9%
  triton_mm_11531 0.4873 ms 81.3%
  triton_mm_11528 0.4940 ms 80.2%
  triton_mm_11520 0.5873 ms 67.5%
  triton_mm_11521 0.6145 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.0694 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0745 ms 100.0%
  triton_mm_8668 0.0757 ms 98.4%
  triton_mm_8662 0.0765 ms 97.4%
  triton_mm_8659 0.0767 ms 97.2%
  triton_mm_8654 0.0772 ms 96.5%
  mm 0.0778 ms 95.8%
  triton_mm_8661 0.0778 ms 95.8%
  triton_mm_8666 0.0782 ms 95.2%
  triton_mm_8665 0.0788 ms 94.5%
  triton_mm_8653 0.0790 ms 94.3%
SingleProcess AUTOTUNE benchmarking takes 1.8877 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8678 0.0321 ms 100.0%
  triton_mm_8680 0.0321 ms 100.0%
  triton_mm_8676 0.0332 ms 96.6%
  triton_mm_8675 0.0332 ms 96.5%
  triton_mm_8677 0.0332 ms 96.5%
  triton_mm_8681 0.0335 ms 95.7%
  triton_mm_8673 0.0351 ms 91.3%
  triton_mm_8684 0.0353 ms 90.8%
  triton_mm_8672 0.0365 ms 87.8%
  triton_mm_8671 0.0366 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.1086 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9686 ms 100.0%
  triton_mm_8702 0.9741 ms 99.4%
  addmm 1.0969 ms 88.3%
  triton_mm_8699 1.1549 ms 83.9%
  triton_mm_8695 1.1781 ms 82.2%
  triton_mm_8703 1.2189 ms 79.5%
  triton_mm_8700 1.2691 ms 76.3%
  triton_mm_8697 1.2806 ms 75.6%
  triton_mm_8696 1.2875 ms 75.2%
  triton_mm_8692 1.5734 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 2.2814 seconds and 0.0021 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3283 ms 100.0%
  mm 0.3389 ms 96.9%
  triton_mm_8714 0.3929 ms 83.5%
  triton_mm_8718 0.3997 ms 82.1%
  triton_mm_8722 0.4027 ms 81.5%
  triton_mm_8715 0.4305 ms 76.2%
  triton_mm_8716 0.4317 ms 76.0%
  triton_mm_8719 0.4360 ms 75.3%
  triton_mm_8711 0.5245 ms 62.6%
  triton_mm_8712 0.5436 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.0590 seconds and 0.0021 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2636 ms 100.0%
  mm 1.2704 ms 99.5%
  triton_mm_8777 1.4903 ms 84.8%
  triton_mm_8769 1.5919 ms 79.4%
  triton_mm_8773 1.6228 ms 77.9%
  triton_mm_8771 1.6685 ms 75.7%
  triton_mm_8770 1.6809 ms 75.2%
  triton_mm_8774 1.6985 ms 74.4%
  triton_mm_8766 2.1217 ms 59.6%
  triton_mm_8767 2.1367 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2379 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9558 0.3295 ms 100.0%
  mm 0.3361 ms 98.1%
  triton_mm_9551 0.3935 ms 83.8%
  triton_mm_9555 0.4000 ms 82.4%
  triton_mm_9559 0.4034 ms 81.7%
  triton_mm_9552 0.4300 ms 76.6%
  triton_mm_9553 0.4320 ms 76.3%
  triton_mm_9556 0.4364 ms 75.5%
  triton_mm_9548 0.5238 ms 62.9%
  triton_mm_9549 0.5429 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0637 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9572 0.0752 ms 100.0%
  triton_mm_9577 0.0758 ms 99.2%
  triton_mm_9568 0.0764 ms 98.4%
  triton_mm_9571 0.0771 ms 97.5%
  triton_mm_9570 0.0776 ms 96.9%
  triton_mm_9563 0.0780 ms 96.3%
  triton_mm_9575 0.0780 ms 96.3%
  mm 0.0783 ms 96.0%
  triton_mm_9574 0.0787 ms 95.5%
  triton_mm_9562 0.0796 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.9044 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9613 1.2638 ms 100.0%
  mm 1.2698 ms 99.5%
  triton_mm_9614 1.4879 ms 84.9%
  triton_mm_9606 1.5983 ms 79.1%
  triton_mm_9610 1.6237 ms 77.8%
  triton_mm_9608 1.6656 ms 75.9%
  triton_mm_9607 1.6817 ms 75.1%
  triton_mm_9611 1.6996 ms 74.4%
  triton_mm_9603 2.1232 ms 59.5%
  triton_mm_9604 2.1364 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2211 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9763 ms 100.0%
  triton_mm_9632 0.9808 ms 99.5%
  addmm 1.1051 ms 88.3%
  triton_mm_9629 1.1668 ms 83.7%
  triton_mm_9625 1.1942 ms 81.7%
  triton_mm_9633 1.2311 ms 79.3%
  triton_mm_9630 1.2696 ms 76.9%
  triton_mm_9626 1.2755 ms 76.5%
  triton_mm_9627 1.2862 ms 75.9%
  triton_mm_9622 1.5936 ms 61.3%
SingleProcess AUTOTUNE benchmarking takes 2.2973 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0387 ms 100.0%
  triton_mm_11428 0.0401 ms 96.6%
  triton_mm_11422 0.0402 ms 96.3%
  triton_mm_11437 0.0404 ms 96.0%
  triton_mm_11424 0.0404 ms 95.8%
  triton_mm_11432 0.0407 ms 95.2%
  triton_mm_11425 0.0409 ms 94.7%
  triton_mm_11435 0.0409 ms 94.6%
  triton_mm_11431 0.0442 ms 87.7%
  mm 0.0451 ms 85.8%
SingleProcess AUTOTUNE benchmarking takes 1.8781 seconds and 0.0058 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.4053 ms 100.0%
  bias_addmm 5.4146 ms 99.8%
  addmm 6.0589 ms 89.2%
  triton_mm_11489 6.5671 ms 82.3%
  triton_mm_11485 6.6935 ms 80.8%
  triton_mm_11493 6.6948 ms 80.7%
  triton_mm_11490 7.1727 ms 75.4%
  triton_mm_11487 7.2513 ms 74.5%
  triton_mm_11486 7.3037 ms 74.0%
  triton_mm_11482 8.8549 ms 61.0%
SingleProcess AUTOTUNE benchmarking takes 2.9541 seconds and 0.0017 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1764 ms 100.0%
  bias_addmm 0.1767 ms 99.9%
  addmm 0.1827 ms 96.6%
  triton_mm_11543 0.1937 ms 91.1%
  triton_mm_11542 0.1939 ms 91.0%
  triton_mm_11544 0.1941 ms 90.9%
  triton_mm_11547 0.1949 ms 90.5%
  triton_mm_11546 0.1955 ms 90.2%
  triton_mm_11550 0.2154 ms 81.9%
  triton_mm_11540 0.2361 ms 74.7%
SingleProcess AUTOTUNE benchmarking takes 2.1095 seconds and 0.0011 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:13, 216.59s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2686 ms 100.0%
  mm 1.4419 ms 88.0%
  triton_mm_8754 1.5620 ms 81.2%
  triton_mm_8758 1.5726 ms 80.7%
  triton_mm_8750 1.6193 ms 78.3%
  triton_mm_8755 1.6834 ms 75.4%
  triton_mm_8752 1.7129 ms 74.1%
  triton_mm_8751 1.7239 ms 73.6%
  triton_mm_8747 2.0833 ms 60.9%
  triton_mm_8748 2.1438 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2160 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2790 ms 100.0%
  mm 1.4423 ms 88.7%
  triton_mm_8847 1.5603 ms 82.0%
  triton_mm_8851 1.5684 ms 81.6%
  triton_mm_8843 1.5890 ms 80.5%
  triton_mm_8848 1.6806 ms 76.1%
  triton_mm_8844 1.7097 ms 74.8%
  triton_mm_8845 1.7189 ms 74.4%
  triton_mm_8840 2.0886 ms 61.2%
  triton_mm_8841 2.1414 ms 59.7%
SingleProcess AUTOTUNE benchmarking takes 2.2364 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8943 1.2819 ms 100.0%
  mm 1.2902 ms 99.4%
  triton_mm_8940 1.5610 ms 82.1%
  triton_mm_8944 1.5693 ms 81.7%
  triton_mm_8936 1.5700 ms 81.6%
  triton_mm_8941 1.6805 ms 76.3%
  triton_mm_8937 1.6973 ms 75.5%
  triton_mm_8938 1.7084 ms 75.0%
  triton_mm_8933 2.1107 ms 60.7%
  triton_mm_8934 2.1619 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2155 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10524 1.2816 ms 100.0%
  mm 1.2905 ms 99.3%
  triton_mm_10521 1.5603 ms 82.1%
  triton_mm_10525 1.5726 ms 81.5%
  triton_mm_10517 1.6073 ms 79.7%
  triton_mm_10522 1.6803 ms 76.3%
  triton_mm_10518 1.6993 ms 75.4%
  triton_mm_10519 1.7210 ms 74.5%
  triton_mm_10514 2.0885 ms 61.4%
  triton_mm_10515 2.1624 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2228 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10896 1.2803 ms 100.0%
  mm 1.2909 ms 99.2%
  triton_mm_10893 1.5560 ms 82.3%
  triton_mm_10889 1.5694 ms 81.6%
  triton_mm_10897 1.5727 ms 81.4%
  triton_mm_10890 1.6790 ms 76.3%
  triton_mm_10894 1.6982 ms 75.4%
  triton_mm_10891 1.7171 ms 74.6%
  triton_mm_10886 2.0925 ms 61.2%
  triton_mm_10887 2.1616 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2230 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11175 1.2793 ms 100.0%
  mm 1.2901 ms 99.2%
  triton_mm_11172 1.5568 ms 82.2%
  triton_mm_11176 1.5708 ms 81.4%
  triton_mm_11168 1.5724 ms 81.4%
  triton_mm_11173 1.7005 ms 75.2%
  triton_mm_11169 1.7011 ms 75.2%
  triton_mm_11170 1.7215 ms 74.3%
  triton_mm_11165 2.1103 ms 60.6%
  triton_mm_11166 2.1667 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2298 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  mm 1.2642 ms 100.0%
  triton_mm_11361 1.2652 ms 99.9%
  triton_mm_11358 1.5599 ms 81.0%
  triton_mm_11362 1.5701 ms 80.5%
  triton_mm_11354 1.5889 ms 79.6%
  triton_mm_11355 1.6996 ms 74.4%
  triton_mm_11359 1.6999 ms 74.4%
  triton_mm_11356 1.7171 ms 73.6%
  triton_mm_11351 2.0924 ms 60.4%
  triton_mm_11352 2.1620 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2265 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3726 ms 100.0%
  mm 0.3856 ms 96.6%
  triton_mm_11504 0.4168 ms 89.4%
  triton_mm_11508 0.4247 ms 87.7%
  triton_mm_11505 0.4484 ms 83.1%
  triton_mm_11512 0.4573 ms 81.5%
  triton_mm_11506 0.4589 ms 81.2%
  triton_mm_11509 0.4714 ms 79.0%
  triton_mm_11501 0.5641 ms 66.1%
  triton_mm_11502 0.5791 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.0533 seconds and 0.0015 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3947 ms 100.0%
  mm 0.4065 ms 97.1%
  triton_mm_11523 0.4343 ms 90.9%
  triton_mm_11527 0.4611 ms 85.6%
  triton_mm_11524 0.4700 ms 84.0%
  triton_mm_11525 0.4836 ms 81.6%
  triton_mm_11531 0.4853 ms 81.3%
  triton_mm_11528 0.4948 ms 79.8%
  triton_mm_11520 0.5916 ms 66.7%
  triton_mm_11521 0.6188 ms 63.8%
SingleProcess AUTOTUNE benchmarking takes 2.0542 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0743 ms 100.0%
  triton_mm_8668 0.0752 ms 98.7%
  triton_mm_8659 0.0754 ms 98.5%
  triton_mm_8662 0.0771 ms 96.4%
  mm 0.0776 ms 95.7%
  triton_mm_8661 0.0776 ms 95.7%
  triton_mm_8654 0.0776 ms 95.7%
  triton_mm_8666 0.0780 ms 95.2%
  triton_mm_8665 0.0781 ms 95.0%
  triton_mm_8653 0.0784 ms 94.8%
SingleProcess AUTOTUNE benchmarking takes 1.8863 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0317 ms 100.0%
  triton_mm_8680 0.0317 ms 99.9%
  triton_mm_8676 0.0318 ms 99.7%
  triton_mm_8677 0.0328 ms 96.6%
  triton_mm_8678 0.0328 ms 96.6%
  triton_mm_8681 0.0331 ms 95.7%
  triton_mm_8673 0.0349 ms 90.8%
  triton_mm_8671 0.0352 ms 90.0%
  triton_mm_8672 0.0361 ms 87.7%
  triton_mm_8684 0.0362 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.1098 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9685 ms 100.0%
  triton_mm_8702 0.9759 ms 99.2%
  addmm 1.1009 ms 88.0%
  triton_mm_8699 1.1653 ms 83.1%
  triton_mm_8695 1.1997 ms 80.7%
  triton_mm_8703 1.2299 ms 78.7%
  triton_mm_8700 1.2820 ms 75.5%
  triton_mm_8696 1.2952 ms 74.8%
  triton_mm_8697 1.2990 ms 74.6%
  triton_mm_8692 1.5716 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 2.2949 seconds and 0.0019 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3257 ms 100.0%
  mm 0.3377 ms 96.4%
  triton_mm_8714 0.3932 ms 82.8%
  triton_mm_8718 0.3956 ms 82.3%
  triton_mm_8722 0.4028 ms 80.9%
  triton_mm_8715 0.4303 ms 75.7%
  triton_mm_8716 0.4330 ms 75.2%
  triton_mm_8719 0.4338 ms 75.1%
  triton_mm_8711 0.5262 ms 61.9%
  triton_mm_8712 0.5447 ms 59.8%
SingleProcess AUTOTUNE benchmarking takes 2.0599 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2689 ms 100.0%
  mm 1.2767 ms 99.4%
  triton_mm_8777 1.5041 ms 84.4%
  triton_mm_8769 1.6128 ms 78.7%
  triton_mm_8773 1.6261 ms 78.0%
  triton_mm_8770 1.6759 ms 75.7%
  triton_mm_8771 1.6800 ms 75.5%
  triton_mm_8774 1.6903 ms 75.1%
  triton_mm_8766 2.1146 ms 60.0%
  triton_mm_8767 2.1589 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.2366 seconds and 0.0015 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8828 0.0742 ms 100.0%
  triton_mm_8833 0.0752 ms 98.7%
  triton_mm_8824 0.0762 ms 97.4%
  triton_mm_8827 0.0770 ms 96.3%
  triton_mm_8819 0.0777 ms 95.5%
  triton_mm_8831 0.0780 ms 95.1%
  triton_mm_8830 0.0780 ms 95.0%
  mm 0.0784 ms 94.7%
  triton_mm_8818 0.0784 ms 94.6%
  triton_mm_8826 0.0785 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.8877 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8869 1.2659 ms 100.0%
  mm 1.2786 ms 99.0%
  triton_mm_8870 1.4828 ms 85.4%
  triton_mm_8862 1.5920 ms 79.5%
  triton_mm_8866 1.6453 ms 76.9%
  triton_mm_8864 1.6781 ms 75.4%
  triton_mm_8867 1.6895 ms 74.9%
  triton_mm_8863 1.6949 ms 74.7%
  triton_mm_8859 2.1133 ms 59.9%
  triton_mm_8860 2.1614 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2281 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8888 0.9757 ms 100.0%
  bias_addmm 0.9772 ms 99.8%
  addmm 1.0989 ms 88.8%
  triton_mm_8885 1.1656 ms 83.7%
  triton_mm_8881 1.1728 ms 83.2%
  triton_mm_8889 1.2310 ms 79.3%
  triton_mm_8886 1.2822 ms 76.1%
  triton_mm_8882 1.2831 ms 76.0%
  triton_mm_8883 1.3000 ms 75.1%
  triton_mm_8878 1.5891 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.2965 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8907 0.3273 ms 100.0%
  mm 0.3374 ms 97.0%
  triton_mm_8900 0.3898 ms 84.0%
  triton_mm_8904 0.3977 ms 82.3%
  triton_mm_8908 0.4035 ms 81.1%
  triton_mm_8901 0.4297 ms 76.2%
  triton_mm_8905 0.4339 ms 75.4%
  triton_mm_8902 0.4340 ms 75.4%
  triton_mm_8897 0.5255 ms 62.3%
  triton_mm_8898 0.5458 ms 60.0%
SingleProcess AUTOTUNE benchmarking takes 2.0595 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9386 0.0750 ms 100.0%
  triton_mm_9391 0.0752 ms 99.8%
  triton_mm_9382 0.0756 ms 99.3%
  triton_mm_9385 0.0763 ms 98.4%
  triton_mm_9377 0.0769 ms 97.5%
  triton_mm_9389 0.0773 ms 97.1%
  triton_mm_9384 0.0775 ms 96.8%
  mm 0.0785 ms 95.6%
  triton_mm_9388 0.0786 ms 95.5%
  triton_mm_9379 0.0790 ms 95.0%
SingleProcess AUTOTUNE benchmarking takes 1.9103 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9427 1.2539 ms 100.0%
  mm 1.2649 ms 99.1%
  triton_mm_9428 1.4886 ms 84.2%
  triton_mm_9420 1.6109 ms 77.8%
  triton_mm_9424 1.6265 ms 77.1%
  triton_mm_9422 1.6804 ms 74.6%
  triton_mm_9421 1.7064 ms 73.5%
  triton_mm_9425 1.7106 ms 73.3%
  triton_mm_9417 2.1100 ms 59.4%
  triton_mm_9418 2.1580 ms 58.1%
SingleProcess AUTOTUNE benchmarking takes 2.2236 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_9446 0.9759 ms 100.0%
  bias_addmm 0.9796 ms 99.6%
  addmm 1.1008 ms 88.7%
  triton_mm_9443 1.1629 ms 83.9%
  triton_mm_9439 1.1735 ms 83.2%
  triton_mm_9447 1.2311 ms 79.3%
  triton_mm_9444 1.2800 ms 76.2%
  triton_mm_9440 1.2834 ms 76.0%
  triton_mm_9441 1.2983 ms 75.2%
  triton_mm_9436 1.5875 ms 61.5%
SingleProcess AUTOTUNE benchmarking takes 2.3042 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9465 0.3286 ms 100.0%
  mm 0.3381 ms 97.2%
  triton_mm_9458 0.3926 ms 83.7%
  triton_mm_9462 0.3967 ms 82.8%
  triton_mm_9466 0.4021 ms 81.7%
  triton_mm_9459 0.4295 ms 76.5%
  triton_mm_9460 0.4343 ms 75.7%
  triton_mm_9463 0.4348 ms 75.6%
  triton_mm_9455 0.5254 ms 62.5%
  triton_mm_9456 0.5459 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 2.0539 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0380 ms 100.0%
  triton_mm_11428 0.0383 ms 99.2%
  triton_mm_11432 0.0391 ms 97.1%
  triton_mm_11422 0.0398 ms 95.3%
  triton_mm_11424 0.0399 ms 95.2%
  triton_mm_11437 0.0400 ms 95.0%
  triton_mm_11425 0.0406 ms 93.5%
  triton_mm_11435 0.0407 ms 93.2%
  triton_mm_11431 0.0431 ms 88.0%
  mm 0.0448 ms 84.8%
SingleProcess AUTOTUNE benchmarking takes 1.8870 seconds and 0.0012 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.4568 ms 100.0%
  bias_addmm 5.4722 ms 99.7%
  addmm 6.0650 ms 90.0%
  triton_mm_11489 6.6594 ms 81.9%
  triton_mm_11485 6.6846 ms 81.6%
  triton_mm_11493 6.7436 ms 80.9%
  triton_mm_11490 7.1652 ms 76.2%
  triton_mm_11486 7.2824 ms 74.9%
  triton_mm_11487 7.2912 ms 74.8%
  triton_mm_11482 8.9552 ms 60.9%
SingleProcess AUTOTUNE benchmarking takes 2.9730 seconds and 0.0020 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1753 ms 100.0%
  bias_addmm 0.1765 ms 99.3%
  addmm 0.1814 ms 96.6%
  triton_mm_11542 0.1907 ms 91.9%
  triton_mm_11543 0.1916 ms 91.5%
  triton_mm_11547 0.1927 ms 90.9%
  triton_mm_11546 0.1940 ms 90.3%
  triton_mm_11544 0.1953 ms 89.7%
  triton_mm_11550 0.2145 ms 81.7%
  triton_mm_11540 0.2345 ms 74.8%
SingleProcess AUTOTUNE benchmarking takes 2.1216 seconds and 0.0014 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:13, 216.59s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2853 ms 100.0%
  mm 1.4433 ms 89.1%
  triton_mm_8754 1.5634 ms 82.2%
  triton_mm_8758 1.5768 ms 81.5%
  triton_mm_8750 1.5899 ms 80.8%
  triton_mm_8755 1.6885 ms 76.1%
  triton_mm_8751 1.7041 ms 75.4%
  triton_mm_8752 1.7275 ms 74.4%
  triton_mm_8747 2.0914 ms 61.5%
  triton_mm_8748 2.1762 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2295 seconds and 0.0023 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2852 ms 100.0%
  mm 1.4430 ms 89.1%
  triton_mm_8847 1.5750 ms 81.6%
  triton_mm_8851 1.5848 ms 81.1%
  triton_mm_8843 1.5910 ms 80.8%
  triton_mm_8848 1.6991 ms 75.6%
  triton_mm_8844 1.7270 ms 74.4%
  triton_mm_8845 1.7362 ms 74.0%
  triton_mm_8840 2.1097 ms 60.9%
  triton_mm_8841 2.1660 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2389 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3733 ms 100.0%
  mm 0.3839 ms 97.2%
  triton_mm_11504 0.4188 ms 89.1%
  triton_mm_11508 0.4272 ms 87.4%
  triton_mm_11505 0.4533 ms 82.4%
  triton_mm_11512 0.4596 ms 81.2%
  triton_mm_11506 0.4666 ms 80.0%
  triton_mm_11509 0.4719 ms 79.1%
  triton_mm_11501 0.5720 ms 65.3%
  triton_mm_11502 0.5865 ms 63.7%
SingleProcess AUTOTUNE benchmarking takes 2.0659 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3957 ms 100.0%
  mm 0.4080 ms 97.0%
  triton_mm_11523 0.4364 ms 90.7%
  triton_mm_11527 0.4567 ms 86.7%
  triton_mm_11524 0.4712 ms 84.0%
  triton_mm_11525 0.4828 ms 82.0%
  triton_mm_11531 0.4857 ms 81.5%
  triton_mm_11528 0.4938 ms 80.1%
  triton_mm_11520 0.5997 ms 66.0%
  triton_mm_11521 0.6186 ms 64.0%
SingleProcess AUTOTUNE benchmarking takes 2.0670 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0746 ms 100.0%
  triton_mm_8668 0.0752 ms 99.1%
  triton_mm_8659 0.0761 ms 98.0%
  triton_mm_8662 0.0767 ms 97.2%
  mm 0.0777 ms 96.0%
  triton_mm_8666 0.0779 ms 95.7%
  triton_mm_8654 0.0781 ms 95.5%
  triton_mm_8661 0.0782 ms 95.3%
  triton_mm_8665 0.0787 ms 94.7%
  triton_mm_8653 0.0790 ms 94.4%
SingleProcess AUTOTUNE benchmarking takes 1.9059 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8678 0.0316 ms 100.0%
  triton_mm_8675 0.0317 ms 99.7%
  triton_mm_8680 0.0326 ms 97.0%
  triton_mm_8677 0.0327 ms 96.7%
  triton_mm_8676 0.0328 ms 96.6%
  triton_mm_8681 0.0328 ms 96.5%
  triton_mm_8673 0.0336 ms 94.2%
  triton_mm_8684 0.0358 ms 88.4%
  triton_mm_8672 0.0358 ms 88.3%
  triton_mm_8671 0.0362 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.1392 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9788 ms 100.0%
  bias_addmm 0.9813 ms 99.7%
  addmm 1.1015 ms 88.9%
  triton_mm_8699 1.1782 ms 83.1%
  triton_mm_8695 1.2024 ms 81.4%
  triton_mm_8703 1.2345 ms 79.3%
  triton_mm_8700 1.2800 ms 76.5%
  triton_mm_8697 1.2964 ms 75.5%
  triton_mm_8696 1.2974 ms 75.4%
  triton_mm_8692 1.5910 ms 61.5%
SingleProcess AUTOTUNE benchmarking takes 2.3177 seconds and 0.0024 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3299 ms 100.0%
  mm 0.3393 ms 97.2%
  triton_mm_8714 0.3956 ms 83.4%
  triton_mm_8718 0.4005 ms 82.4%
  triton_mm_8722 0.4031 ms 81.8%
  triton_mm_8715 0.4328 ms 76.2%
  triton_mm_8719 0.4397 ms 75.0%
  triton_mm_8716 0.4403 ms 74.9%
  triton_mm_8711 0.5316 ms 62.1%
  triton_mm_8712 0.5541 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.0616 seconds and 0.0021 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2656 ms 100.0%
  mm 1.2812 ms 98.8%
  triton_mm_8777 1.4996 ms 84.4%
  triton_mm_8769 1.6322 ms 77.5%
  triton_mm_8773 1.6532 ms 76.6%
  triton_mm_8770 1.6949 ms 74.7%
  triton_mm_8771 1.6991 ms 74.5%
  triton_mm_8774 1.7128 ms 73.9%
  triton_mm_8766 2.1473 ms 58.9%
  triton_mm_8767 2.1622 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2414 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0390 ms 100.0%
  triton_mm_11424 0.0390 ms 100.0%
  triton_mm_11428 0.0393 ms 99.2%
  triton_mm_11437 0.0398 ms 97.9%
  triton_mm_11432 0.0404 ms 96.6%
  triton_mm_11422 0.0404 ms 96.6%
  triton_mm_11435 0.0409 ms 95.4%
  triton_mm_11425 0.0413 ms 94.5%
  triton_mm_11431 0.0440 ms 88.6%
  mm 0.0449 ms 86.9%
SingleProcess AUTOTUNE benchmarking takes 1.8891 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.4641 ms 100.0%
  bias_addmm 5.4798 ms 99.7%
  addmm 6.1275 ms 89.2%
  triton_mm_11489 6.6692 ms 81.9%
  triton_mm_11493 6.7575 ms 80.9%
  triton_mm_11485 6.8666 ms 79.6%
  triton_mm_11490 7.2652 ms 75.2%
  triton_mm_11486 7.2991 ms 74.9%
  triton_mm_11487 7.3037 ms 74.8%
  triton_mm_11482 8.9731 ms 60.9%
SingleProcess AUTOTUNE benchmarking takes 2.9935 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1753 ms 100.0%
  bias_addmm 0.1769 ms 99.1%
  addmm 0.1825 ms 96.1%
  triton_mm_11543 0.1923 ms 91.2%
  triton_mm_11542 0.1924 ms 91.1%
  triton_mm_11544 0.1939 ms 90.4%
  triton_mm_11547 0.1941 ms 90.3%
  triton_mm_11546 0.1946 ms 90.1%
  triton_mm_11550 0.2146 ms 81.7%
  triton_mm_11540 0.2342 ms 74.9%
SingleProcess AUTOTUNE benchmarking takes 2.1208 seconds and 0.0012 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:13, 216.59s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2675 ms 100.0%
  mm 1.4393 ms 88.1%
  triton_mm_8754 1.5416 ms 82.2%
  triton_mm_8750 1.5698 ms 80.7%
  triton_mm_8758 1.5722 ms 80.6%
  triton_mm_8755 1.6832 ms 75.3%
  triton_mm_8751 1.7032 ms 74.4%
  triton_mm_8752 1.7093 ms 74.2%
  triton_mm_8747 2.0849 ms 60.8%
  triton_mm_8748 2.1440 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2304 seconds and 0.0020 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2670 ms 100.0%
  mm 1.4425 ms 87.8%
  triton_mm_8847 1.5551 ms 81.5%
  triton_mm_8851 1.5700 ms 80.7%
  triton_mm_8843 1.6303 ms 77.7%
  triton_mm_8848 1.6786 ms 75.5%
  triton_mm_8844 1.6999 ms 74.5%
  triton_mm_8845 1.7009 ms 74.5%
  triton_mm_8840 2.0782 ms 61.0%
  triton_mm_8841 2.1390 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2376 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3736 ms 100.0%
  mm 0.3817 ms 97.9%
  triton_mm_11504 0.4154 ms 89.9%
  triton_mm_11508 0.4235 ms 88.2%
  triton_mm_11505 0.4442 ms 84.1%
  triton_mm_11512 0.4582 ms 81.5%
  triton_mm_11506 0.4590 ms 81.4%
  triton_mm_11509 0.4675 ms 79.9%
  triton_mm_11501 0.5609 ms 66.6%
  triton_mm_11502 0.5817 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.0576 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3943 ms 100.0%
  mm 0.4053 ms 97.3%
  triton_mm_11523 0.4380 ms 90.0%
  triton_mm_11527 0.4569 ms 86.3%
  triton_mm_11524 0.4707 ms 83.8%
  triton_mm_11525 0.4832 ms 81.6%
  triton_mm_11531 0.4863 ms 81.1%
  triton_mm_11528 0.4911 ms 80.3%
  triton_mm_11520 0.5943 ms 66.4%
  triton_mm_11521 0.6142 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.0660 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0742 ms 100.0%
  triton_mm_8668 0.0752 ms 98.6%
  triton_mm_8659 0.0761 ms 97.5%
  triton_mm_8662 0.0770 ms 96.3%
  triton_mm_8666 0.0774 ms 95.8%
  triton_mm_8661 0.0775 ms 95.8%
  triton_mm_8654 0.0779 ms 95.3%
  triton_mm_8653 0.0781 ms 95.0%
  triton_mm_8665 0.0781 ms 95.0%
  mm 0.0785 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.8860 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8678 0.0329 ms 100.0%
  triton_mm_8675 0.0329 ms 99.9%
  triton_mm_8677 0.0329 ms 99.9%
  triton_mm_8680 0.0330 ms 99.8%
  triton_mm_8676 0.0330 ms 99.7%
  triton_mm_8681 0.0332 ms 99.0%
  triton_mm_8673 0.0338 ms 97.3%
  triton_mm_8672 0.0351 ms 93.6%
  triton_mm_8684 0.0363 ms 90.7%
  triton_mm_8671 0.0364 ms 90.5%
SingleProcess AUTOTUNE benchmarking takes 2.1077 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9639 ms 100.0%
  triton_mm_8702 0.9777 ms 98.6%
  addmm 1.1009 ms 87.6%
  triton_mm_8699 1.1639 ms 82.8%
  triton_mm_8695 1.1731 ms 82.2%
  triton_mm_8703 1.2297 ms 78.4%
  triton_mm_8700 1.2675 ms 76.0%
  triton_mm_8697 1.2904 ms 74.7%
  triton_mm_8696 1.2969 ms 74.3%
  triton_mm_8692 1.5711 ms 61.3%
SingleProcess AUTOTUNE benchmarking takes 2.2947 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3286 ms 100.0%
  mm 0.3384 ms 97.1%
  triton_mm_8714 0.3941 ms 83.4%
  triton_mm_8718 0.3995 ms 82.3%
  triton_mm_8722 0.4031 ms 81.5%
  triton_mm_8715 0.4286 ms 76.7%
  triton_mm_8719 0.4347 ms 75.6%
  triton_mm_8716 0.4352 ms 75.5%
  triton_mm_8711 0.5244 ms 62.7%
  triton_mm_8712 0.5436 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0498 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2584 ms 100.0%
  mm 1.2667 ms 99.3%
  triton_mm_8777 1.4861 ms 84.7%
  triton_mm_8769 1.5985 ms 78.7%
  triton_mm_8773 1.6335 ms 77.0%
  triton_mm_8770 1.6762 ms 75.1%
  triton_mm_8771 1.6866 ms 74.6%
  triton_mm_8774 1.6930 ms 74.3%
  triton_mm_8766 2.1147 ms 59.5%
  triton_mm_8767 2.1567 ms 58.3%
SingleProcess AUTOTUNE benchmarking takes 2.2343 seconds and 0.0019 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  mm 1.2667 ms 100.0%
  triton_mm_9148 1.2680 ms 99.9%
  triton_mm_9149 1.4996 ms 84.5%
  triton_mm_9141 1.5975 ms 79.3%
  triton_mm_9145 1.6348 ms 77.5%
  triton_mm_9143 1.6793 ms 75.4%
  triton_mm_9146 1.6930 ms 74.8%
  triton_mm_9142 1.6951 ms 74.7%
  triton_mm_9138 2.1210 ms 59.7%
  triton_mm_9139 2.1571 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2369 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9730 ms 100.0%
  triton_mm_9167 0.9766 ms 99.6%
  addmm 1.1019 ms 88.3%
  triton_mm_9164 1.1626 ms 83.7%
  triton_mm_9160 1.1709 ms 83.1%
  triton_mm_9168 1.2280 ms 79.2%
  triton_mm_9165 1.2659 ms 76.9%
  triton_mm_9162 1.2863 ms 75.6%
  triton_mm_9161 1.2950 ms 75.1%
  triton_mm_9157 1.5849 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.2879 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9186 0.3272 ms 100.0%
  mm 0.3394 ms 96.4%
  triton_mm_9179 0.3944 ms 83.0%
  triton_mm_9183 0.3987 ms 82.1%
  triton_mm_9187 0.4025 ms 81.3%
  triton_mm_9180 0.4302 ms 76.1%
  triton_mm_9181 0.4321 ms 75.7%
  triton_mm_9184 0.4344 ms 75.3%
  triton_mm_9176 0.5247 ms 62.4%
  triton_mm_9177 0.5433 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 2.0502 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_10860 0.3279 ms 100.0%
  mm 0.3363 ms 97.5%
  triton_mm_10853 0.3922 ms 83.6%
  triton_mm_10857 0.3997 ms 82.1%
  triton_mm_10861 0.4030 ms 81.4%
  triton_mm_10854 0.4297 ms 76.3%
  triton_mm_10858 0.4347 ms 75.4%
  triton_mm_10855 0.4352 ms 75.4%
  triton_mm_10850 0.5220 ms 62.8%
  triton_mm_10851 0.5460 ms 60.1%
SingleProcess AUTOTUNE benchmarking takes 2.0473 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0393 ms 100.0%
  triton_mm_11432 0.0394 ms 99.7%
  triton_mm_11422 0.0395 ms 99.4%
  triton_mm_11428 0.0395 ms 99.4%
  triton_mm_11424 0.0396 ms 99.2%
  triton_mm_11437 0.0400 ms 98.1%
  triton_mm_11435 0.0410 ms 95.9%
  triton_mm_11425 0.0415 ms 94.7%
  triton_mm_11431 0.0433 ms 90.7%
  mm 0.0450 ms 87.2%
SingleProcess AUTOTUNE benchmarking takes 1.8599 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3914 ms 100.0%
  bias_addmm 5.3993 ms 99.9%
  addmm 6.0457 ms 89.2%
  triton_mm_11489 6.5755 ms 82.0%
  triton_mm_11493 6.6734 ms 80.8%
  triton_mm_11485 6.7659 ms 79.7%
  triton_mm_11490 7.1738 ms 75.2%
  triton_mm_11487 7.2662 ms 74.2%
  triton_mm_11486 7.2762 ms 74.1%
  triton_mm_11482 8.8328 ms 61.0%
SingleProcess AUTOTUNE benchmarking takes 2.9671 seconds and 0.0014 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1752 ms 100.0%
  bias_addmm 0.1775 ms 98.7%
  addmm 0.1823 ms 96.1%
  triton_mm_11543 0.1921 ms 91.2%
  triton_mm_11547 0.1934 ms 90.6%
  triton_mm_11544 0.1943 ms 90.2%
  triton_mm_11542 0.1953 ms 89.7%
  triton_mm_11546 0.1985 ms 88.3%
  triton_mm_11550 0.2140 ms 81.9%
  triton_mm_11540 0.2349 ms 74.6%
SingleProcess AUTOTUNE benchmarking takes 2.1099 seconds and 0.0011 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:13, 216.59s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2822 ms 100.0%
  mm 1.4413 ms 89.0%
  triton_mm_8754 1.5590 ms 82.2%
  triton_mm_8750 1.5716 ms 81.6%
  triton_mm_8758 1.5742 ms 81.5%
  triton_mm_8751 1.6848 ms 76.1%
  triton_mm_8755 1.7006 ms 75.4%
  triton_mm_8752 1.7213 ms 74.5%
  triton_mm_8747 2.0863 ms 61.5%
  triton_mm_8748 2.1699 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2076 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2818 ms 100.0%
  mm 1.4432 ms 88.8%
  triton_mm_8847 1.5580 ms 82.3%
  triton_mm_8851 1.5720 ms 81.5%
  triton_mm_8843 1.5732 ms 81.5%
  triton_mm_8844 1.7001 ms 75.4%
  triton_mm_8848 1.7022 ms 75.3%
  triton_mm_8845 1.7189 ms 74.6%
  triton_mm_8840 2.0862 ms 61.4%
  triton_mm_8841 2.1641 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2067 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8943 1.2822 ms 100.0%
  mm 1.3092 ms 97.9%
  triton_mm_8940 1.5590 ms 82.2%
  triton_mm_8944 1.5720 ms 81.6%
  triton_mm_8936 1.5766 ms 81.3%
  triton_mm_8937 1.6996 ms 75.4%
  triton_mm_8941 1.7026 ms 75.3%
  triton_mm_8938 1.7174 ms 74.7%
  triton_mm_8933 2.1051 ms 60.9%
  triton_mm_8934 2.1639 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2065 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9408 1.2725 ms 100.0%
  mm 1.3088 ms 97.2%
  triton_mm_9405 1.5581 ms 81.7%
  triton_mm_9409 1.5739 ms 80.9%
  triton_mm_9401 1.5740 ms 80.8%
  triton_mm_9402 1.6995 ms 74.9%
  triton_mm_9406 1.7021 ms 74.8%
  triton_mm_9403 1.7168 ms 74.1%
  triton_mm_9398 2.1057 ms 60.4%
  triton_mm_9399 2.1639 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.2134 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10059 1.2820 ms 100.0%
  mm 1.3082 ms 98.0%
  triton_mm_10056 1.5588 ms 82.2%
  triton_mm_10060 1.5721 ms 81.5%
  triton_mm_10052 1.5893 ms 80.7%
  triton_mm_10053 1.6994 ms 75.4%
  triton_mm_10057 1.7022 ms 75.3%
  triton_mm_10054 1.7175 ms 74.6%
  triton_mm_10049 2.1046 ms 60.9%
  triton_mm_10050 2.1638 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2073 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10431 1.2817 ms 100.0%
  mm 1.3077 ms 98.0%
  triton_mm_10428 1.5588 ms 82.2%
  triton_mm_10432 1.5717 ms 81.6%
  triton_mm_10424 1.5908 ms 80.6%
  triton_mm_10425 1.6993 ms 75.4%
  triton_mm_10429 1.7023 ms 75.3%
  triton_mm_10426 1.7179 ms 74.6%
  triton_mm_10421 2.1048 ms 60.9%
  triton_mm_10422 2.1634 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2075 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10524 1.2746 ms 100.0%
  mm 1.3138 ms 97.0%
  triton_mm_10521 1.5582 ms 81.8%
  triton_mm_10525 1.5824 ms 80.5%
  triton_mm_10517 1.5903 ms 80.1%
  triton_mm_10518 1.6995 ms 75.0%
  triton_mm_10522 1.7023 ms 74.9%
  triton_mm_10519 1.7195 ms 74.1%
  triton_mm_10514 2.1060 ms 60.5%
  triton_mm_10515 2.1638 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.2101 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11361 1.2920 ms 100.0%
  mm 1.3078 ms 98.8%
  triton_mm_11358 1.5578 ms 82.9%
  triton_mm_11354 1.5732 ms 82.1%
  triton_mm_11362 1.5830 ms 81.6%
  triton_mm_11355 1.7001 ms 76.0%
  triton_mm_11359 1.7022 ms 75.9%
  triton_mm_11356 1.7147 ms 75.3%
  triton_mm_11351 2.1046 ms 61.4%
  triton_mm_11352 2.1642 ms 59.7%
SingleProcess AUTOTUNE benchmarking takes 2.2189 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11454 1.2879 ms 100.0%
  mm 1.3118 ms 98.2%
  triton_mm_11451 1.5577 ms 82.7%
  triton_mm_11455 1.5948 ms 80.8%
  triton_mm_11447 1.6094 ms 80.0%
  triton_mm_11452 1.6999 ms 75.8%
  triton_mm_11449 1.7146 ms 75.1%
  triton_mm_11448 1.7168 ms 75.0%
  triton_mm_11444 2.1026 ms 61.2%
  triton_mm_11445 2.1614 ms 59.6%
SingleProcess AUTOTUNE benchmarking takes 2.2164 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3792 ms 100.0%
  mm 0.3891 ms 97.5%
  triton_mm_11504 0.4215 ms 90.0%
  triton_mm_11508 0.4283 ms 88.5%
  triton_mm_11505 0.4523 ms 83.8%
  triton_mm_11512 0.4607 ms 82.3%
  triton_mm_11506 0.4631 ms 81.9%
  triton_mm_11509 0.4718 ms 80.4%
  triton_mm_11501 0.5687 ms 66.7%
  triton_mm_11502 0.5860 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 2.0583 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3999 ms 100.0%
  mm 0.4110 ms 97.3%
  triton_mm_11523 0.4420 ms 90.5%
  triton_mm_11527 0.4635 ms 86.3%
  triton_mm_11524 0.4752 ms 84.1%
  triton_mm_11525 0.4853 ms 82.4%
  triton_mm_11531 0.4896 ms 81.7%
  triton_mm_11528 0.4991 ms 80.1%
  triton_mm_11520 0.5947 ms 67.2%
  triton_mm_11521 0.6204 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.0540 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0742 ms 100.0%
  triton_mm_8659 0.0754 ms 98.5%
  triton_mm_8668 0.0763 ms 97.3%
  triton_mm_8654 0.0767 ms 96.7%
  triton_mm_8666 0.0769 ms 96.5%
  triton_mm_8662 0.0770 ms 96.5%
  triton_mm_8661 0.0771 ms 96.3%
  mm 0.0776 ms 95.7%
  triton_mm_8653 0.0784 ms 94.7%
  triton_mm_8665 0.0784 ms 94.7%
SingleProcess AUTOTUNE benchmarking takes 1.8802 seconds and 0.0017 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0318 ms 100.0%
  triton_mm_8681 0.0320 ms 99.4%
  triton_mm_8676 0.0331 ms 96.0%
  triton_mm_8677 0.0331 ms 96.0%
  triton_mm_8678 0.0331 ms 96.0%
  triton_mm_8680 0.0331 ms 95.9%
  triton_mm_8673 0.0350 ms 90.9%
  triton_mm_8671 0.0353 ms 90.1%
  triton_mm_8672 0.0362 ms 87.7%
  triton_mm_8684 0.0363 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.1045 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9798 ms 100.0%
  triton_mm_8702 0.9834 ms 99.6%
  addmm 1.1079 ms 88.4%
  triton_mm_8699 1.1664 ms 84.0%
  triton_mm_8695 1.1794 ms 83.1%
  triton_mm_8703 1.2304 ms 79.6%
  triton_mm_8700 1.2817 ms 76.4%
  triton_mm_8696 1.2848 ms 76.3%
  triton_mm_8697 1.2912 ms 75.9%
  triton_mm_8692 1.5859 ms 61.8%
SingleProcess AUTOTUNE benchmarking takes 2.2997 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3336 ms 100.0%
  mm 0.3432 ms 97.2%
  triton_mm_8714 0.3977 ms 83.9%
  triton_mm_8718 0.4038 ms 82.6%
  triton_mm_8722 0.4060 ms 82.2%
  triton_mm_8716 0.4348 ms 76.7%
  triton_mm_8715 0.4349 ms 76.7%
  triton_mm_8719 0.4402 ms 75.8%
  triton_mm_8711 0.5300 ms 62.9%
  triton_mm_8712 0.5512 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0584 seconds and 0.0020 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2775 ms 100.0%
  mm 1.2832 ms 99.5%
  triton_mm_8777 1.4997 ms 85.2%
  triton_mm_8769 1.6124 ms 79.2%
  triton_mm_8773 1.6463 ms 77.6%
  triton_mm_8771 1.6700 ms 76.5%
  triton_mm_8770 1.6972 ms 75.3%
  triton_mm_8774 1.7115 ms 74.6%
  triton_mm_8766 2.1444 ms 59.6%
  triton_mm_8767 2.1616 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2250 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9520 1.2724 ms 100.0%
  mm 1.2844 ms 99.1%
  triton_mm_9521 1.5025 ms 84.7%
  triton_mm_9513 1.6127 ms 78.9%
  triton_mm_9517 1.6479 ms 77.2%
  triton_mm_9515 1.6752 ms 76.0%
  triton_mm_9514 1.6968 ms 75.0%
  triton_mm_9518 1.7126 ms 74.3%
  triton_mm_9510 2.1417 ms 59.4%
  triton_mm_9511 2.1620 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.2285 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_9539 0.9834 ms 100.0%
  bias_addmm 0.9878 ms 99.6%
  addmm 1.1171 ms 88.0%
  triton_mm_9536 1.1680 ms 84.2%
  triton_mm_9532 1.1870 ms 82.9%
  triton_mm_9540 1.2340 ms 79.7%
  triton_mm_9537 1.2814 ms 76.7%
  triton_mm_9533 1.2987 ms 75.7%
  triton_mm_9534 1.3053 ms 75.3%
  triton_mm_9529 1.5865 ms 62.0%
SingleProcess AUTOTUNE benchmarking takes 2.3015 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9558 0.3322 ms 100.0%
  mm 0.3440 ms 96.6%
  triton_mm_9551 0.3992 ms 83.2%
  triton_mm_9555 0.4033 ms 82.4%
  triton_mm_9559 0.4070 ms 81.6%
  triton_mm_9553 0.4345 ms 76.5%
  triton_mm_9552 0.4360 ms 76.2%
  triton_mm_9556 0.4415 ms 75.3%
  triton_mm_9548 0.5310 ms 62.6%
  triton_mm_9549 0.5494 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0477 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9572 0.0751 ms 100.0%
  triton_mm_9568 0.0756 ms 99.4%
  triton_mm_9577 0.0756 ms 99.4%
  triton_mm_9571 0.0763 ms 98.5%
  triton_mm_9575 0.0771 ms 97.5%
  triton_mm_9570 0.0776 ms 96.9%
  triton_mm_9574 0.0776 ms 96.8%
  triton_mm_9563 0.0780 ms 96.4%
  mm 0.0784 ms 95.8%
  triton_mm_9562 0.0785 ms 95.7%
SingleProcess AUTOTUNE benchmarking takes 1.8873 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0393 ms 100.0%
  triton_mm_11432 0.0393 ms 99.8%
  triton_mm_11428 0.0397 ms 98.8%
  triton_mm_11424 0.0401 ms 98.0%
  triton_mm_11422 0.0402 ms 97.8%
  triton_mm_11437 0.0403 ms 97.5%
  triton_mm_11435 0.0407 ms 96.4%
  triton_mm_11425 0.0415 ms 94.5%
  triton_mm_11431 0.0429 ms 91.5%
  mm 0.0449 ms 87.4%
SingleProcess AUTOTUNE benchmarking takes 1.8827 seconds and 0.0013 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  bias_addmm 5.4033 ms 100.0%
  triton_mm_11492 5.4184 ms 99.7%
  addmm 6.1176 ms 88.3%
  triton_mm_11489 6.6433 ms 81.3%
  triton_mm_11485 6.6714 ms 81.0%
  triton_mm_11493 6.7101 ms 80.5%
  triton_mm_11490 7.1800 ms 75.3%
  triton_mm_11487 7.2794 ms 74.2%
  triton_mm_11486 7.2825 ms 74.2%
  triton_mm_11482 8.9361 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.9683 seconds and 0.0012 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1762 ms 100.0%
  bias_addmm 0.1783 ms 98.8%
  addmm 0.1844 ms 95.6%
  triton_mm_11542 0.1937 ms 90.9%
  triton_mm_11543 0.1938 ms 90.9%
  triton_mm_11544 0.1952 ms 90.3%
  triton_mm_11546 0.1964 ms 89.7%
  triton_mm_11547 0.1965 ms 89.6%
  triton_mm_11550 0.2144 ms 82.2%
  triton_mm_11540 0.2379 ms 74.1%
SingleProcess AUTOTUNE benchmarking takes 2.1221 seconds and 0.0011 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:13, 216.59s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2408 ms 100.0%
  mm 1.4496 ms 85.6%
  triton_mm_8754 1.5393 ms 80.6%
  triton_mm_8758 1.5537 ms 79.9%
  triton_mm_8750 1.5657 ms 79.2%
  triton_mm_8755 1.6617 ms 74.7%
  triton_mm_8751 1.6648 ms 74.5%
  triton_mm_8752 1.6889 ms 73.5%
  triton_mm_8747 2.0907 ms 59.3%
  triton_mm_8748 2.1505 ms 57.7%
SingleProcess AUTOTUNE benchmarking takes 2.2086 seconds and 0.0018 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3689 ms 100.0%
  triton_mm_11504 0.4100 ms 90.0%
  triton_mm_11508 0.4203 ms 87.8%
  mm 0.4336 ms 85.1%
  triton_mm_11505 0.4394 ms 84.0%
  triton_mm_11506 0.4517 ms 81.7%
  triton_mm_11512 0.4537 ms 81.3%
  triton_mm_11509 0.4621 ms 79.8%
  triton_mm_11501 0.5593 ms 66.0%
  triton_mm_11502 0.5683 ms 64.9%
SingleProcess AUTOTUNE benchmarking takes 2.0470 seconds and 0.0015 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3891 ms 100.0%
  mm 0.4055 ms 95.9%
  triton_mm_11523 0.4318 ms 90.1%
  triton_mm_11527 0.4546 ms 85.6%
  triton_mm_11524 0.4663 ms 83.4%
  triton_mm_11525 0.4728 ms 82.3%
  triton_mm_11531 0.4800 ms 81.1%
  triton_mm_11528 0.4881 ms 79.7%
  triton_mm_11520 0.5816 ms 66.9%
  triton_mm_11521 0.6055 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.0556 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0743 ms 100.0%
  triton_mm_8668 0.0759 ms 97.9%
  triton_mm_8659 0.0761 ms 97.6%
  triton_mm_8662 0.0769 ms 96.6%
  triton_mm_8666 0.0773 ms 96.1%
  triton_mm_8654 0.0776 ms 95.8%
  triton_mm_8653 0.0779 ms 95.4%
  triton_mm_8665 0.0780 ms 95.3%
  triton_mm_8661 0.0782 ms 94.9%
  mm 0.0784 ms 94.8%
SingleProcess AUTOTUNE benchmarking takes 1.8967 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0317 ms 100.0%
  triton_mm_8677 0.0327 ms 96.9%
  triton_mm_8676 0.0328 ms 96.8%
  triton_mm_8678 0.0329 ms 96.5%
  triton_mm_8680 0.0329 ms 96.4%
  triton_mm_8681 0.0331 ms 95.9%
  triton_mm_8673 0.0337 ms 94.0%
  triton_mm_8671 0.0352 ms 90.2%
  triton_mm_8672 0.0360 ms 88.1%
  triton_mm_8684 0.0360 ms 88.0%
SingleProcess AUTOTUNE benchmarking takes 2.1127 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9677 ms 100.0%
  bias_addmm 0.9740 ms 99.4%
  addmm 1.1030 ms 87.7%
  triton_mm_8699 1.1527 ms 84.0%
  triton_mm_8695 1.1746 ms 82.4%
  triton_mm_8703 1.2047 ms 80.3%
  triton_mm_8700 1.2661 ms 76.4%
  triton_mm_8697 1.2725 ms 76.0%
  triton_mm_8696 1.2809 ms 75.6%
  triton_mm_8692 1.5565 ms 62.2%
SingleProcess AUTOTUNE benchmarking takes 2.2888 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3255 ms 100.0%
  mm 0.3357 ms 97.0%
  triton_mm_8714 0.3904 ms 83.4%
  triton_mm_8718 0.3957 ms 82.2%
  triton_mm_8722 0.4029 ms 80.8%
  triton_mm_8715 0.4260 ms 76.4%
  triton_mm_8716 0.4292 ms 75.8%
  triton_mm_8719 0.4309 ms 75.5%
  triton_mm_8711 0.5216 ms 62.4%
  triton_mm_8712 0.5385 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.0432 seconds and 0.0017 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2548 ms 100.0%
  mm 1.2661 ms 99.1%
  triton_mm_8777 1.4695 ms 85.4%
  triton_mm_8769 1.5834 ms 79.2%
  triton_mm_8773 1.6163 ms 77.6%
  triton_mm_8770 1.6602 ms 75.6%
  triton_mm_8771 1.6731 ms 75.0%
  triton_mm_8774 1.6889 ms 74.3%
  triton_mm_8766 2.1001 ms 59.7%
  triton_mm_8767 2.1364 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2194 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9334 1.2470 ms 100.0%
  mm 1.2555 ms 99.3%
  triton_mm_9335 1.4802 ms 84.3%
  triton_mm_9327 1.6004 ms 77.9%
  triton_mm_9331 1.6171 ms 77.1%
  triton_mm_9332 1.6718 ms 74.6%
  triton_mm_9328 1.6816 ms 74.2%
  triton_mm_9329 1.6820 ms 74.1%
  triton_mm_9324 2.1132 ms 59.0%
  triton_mm_9325 2.1452 ms 58.1%
SingleProcess AUTOTUNE benchmarking takes 2.2165 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9372 0.3251 ms 100.0%
  mm 0.3350 ms 97.0%
  triton_mm_9365 0.3902 ms 83.3%
  triton_mm_9369 0.3985 ms 81.6%
  triton_mm_9373 0.4016 ms 80.9%
  triton_mm_9366 0.4265 ms 76.2%
  triton_mm_9367 0.4290 ms 75.8%
  triton_mm_9370 0.4322 ms 75.2%
  triton_mm_9362 0.5212 ms 62.4%
  triton_mm_9363 0.5402 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 2.0424 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_10953 0.3249 ms 100.0%
  mm 0.3329 ms 97.6%
  triton_mm_10946 0.3915 ms 83.0%
  triton_mm_10950 0.3981 ms 81.6%
  triton_mm_10954 0.4019 ms 80.9%
  triton_mm_10947 0.4265 ms 76.2%
  triton_mm_10948 0.4303 ms 75.5%
  triton_mm_10951 0.4330 ms 75.0%
  triton_mm_10943 0.5212 ms 62.3%
  triton_mm_10944 0.5386 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.0412 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_11008 1.2556 ms 100.0%
  mm 1.2650 ms 99.3%
  triton_mm_11009 1.4696 ms 85.4%
  triton_mm_11001 1.5940 ms 78.8%
  triton_mm_11005 1.6194 ms 77.5%
  triton_mm_11006 1.6726 ms 75.1%
  triton_mm_11003 1.6756 ms 74.9%
  triton_mm_11002 1.6965 ms 74.0%
  triton_mm_10998 2.0987 ms 59.8%
  triton_mm_10999 2.1019 ms 59.7%
SingleProcess AUTOTUNE benchmarking takes 2.2101 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_11101 1.2555 ms 100.0%
  mm 1.2650 ms 99.3%
  triton_mm_11102 1.4710 ms 85.3%
  triton_mm_11094 1.5876 ms 79.1%
  triton_mm_11098 1.6357 ms 76.8%
  triton_mm_11096 1.6787 ms 74.8%
  triton_mm_11095 1.6942 ms 74.1%
  triton_mm_11099 1.6960 ms 74.0%
  triton_mm_11091 2.1010 ms 59.8%
  triton_mm_11092 2.1373 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2203 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11424 0.0389 ms 100.0%
  triton_mm_11423 0.0390 ms 99.7%
  triton_mm_11428 0.0394 ms 98.7%
  triton_mm_11437 0.0399 ms 97.4%
  triton_mm_11422 0.0401 ms 97.0%
  triton_mm_11432 0.0404 ms 96.4%
  triton_mm_11435 0.0409 ms 95.2%
  triton_mm_11425 0.0409 ms 95.1%
  triton_mm_11431 0.0441 ms 88.3%
  mm 0.0450 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 1.8607 seconds and 0.0012 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3264 ms 100.0%
  bias_addmm 5.3572 ms 99.4%
  addmm 6.0419 ms 88.2%
  triton_mm_11489 6.5812 ms 80.9%
  triton_mm_11493 6.5996 ms 80.7%
  triton_mm_11485 6.6870 ms 79.7%
  triton_mm_11490 7.0946 ms 75.1%
  triton_mm_11486 7.1963 ms 74.0%
  triton_mm_11487 7.2109 ms 73.9%
  triton_mm_11482 8.8508 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 2.9627 seconds and 0.0013 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1743 ms 100.0%
  bias_addmm 0.1754 ms 99.4%
  addmm 0.1810 ms 96.3%
  triton_mm_11543 0.1904 ms 91.6%
  triton_mm_11547 0.1923 ms 90.6%
  triton_mm_11542 0.1931 ms 90.3%
  triton_mm_11544 0.1939 ms 89.9%
  triton_mm_11546 0.1960 ms 88.9%
  triton_mm_11550 0.2138 ms 81.5%
  triton_mm_11540 0.2353 ms 74.1%
SingleProcess AUTOTUNE benchmarking takes 2.1158 seconds and 0.0013 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:13, 216.60s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2671 ms 100.0%
  mm 1.4457 ms 87.6%
  triton_mm_8754 1.5564 ms 81.4%
  triton_mm_8758 1.5716 ms 80.6%
  triton_mm_8750 1.5746 ms 80.5%
  triton_mm_8755 1.6803 ms 75.4%
  triton_mm_8751 1.6996 ms 74.5%
  triton_mm_8752 1.7064 ms 74.3%
  triton_mm_8747 2.0839 ms 60.8%
  triton_mm_8748 2.1653 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2076 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8850 1.2810 ms 100.0%
  mm 1.4435 ms 88.7%
  triton_mm_8847 1.5546 ms 82.4%
  triton_mm_8851 1.5692 ms 81.6%
  triton_mm_8843 1.5718 ms 81.5%
  triton_mm_8848 1.6817 ms 76.2%
  triton_mm_8844 1.6842 ms 76.1%
  triton_mm_8845 1.7180 ms 74.6%
  triton_mm_8840 2.0809 ms 61.6%
  triton_mm_8841 2.1627 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2097 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11082 1.2686 ms 100.0%
  mm 1.2987 ms 97.7%
  triton_mm_11079 1.5545 ms 81.6%
  triton_mm_11083 1.5689 ms 80.9%
  triton_mm_11075 1.6086 ms 78.9%
  triton_mm_11080 1.6806 ms 75.5%
  triton_mm_11076 1.7011 ms 74.6%
  triton_mm_11077 1.7060 ms 74.4%
  triton_mm_11072 2.1032 ms 60.3%
  triton_mm_11073 2.1630 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2164 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3774 ms 100.0%
  mm 0.3863 ms 97.7%
  triton_mm_11504 0.4191 ms 90.1%
  triton_mm_11508 0.4273 ms 88.3%
  triton_mm_11505 0.4505 ms 83.8%
  triton_mm_11512 0.4600 ms 82.0%
  triton_mm_11506 0.4613 ms 81.8%
  triton_mm_11509 0.4708 ms 80.2%
  triton_mm_11501 0.5640 ms 66.9%
  triton_mm_11502 0.5862 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.0539 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3978 ms 100.0%
  mm 0.4082 ms 97.5%
  triton_mm_11523 0.4378 ms 90.9%
  triton_mm_11527 0.4618 ms 86.2%
  triton_mm_11524 0.4738 ms 84.0%
  triton_mm_11525 0.4802 ms 82.8%
  triton_mm_11531 0.4882 ms 81.5%
  triton_mm_11528 0.4964 ms 80.1%
  triton_mm_11520 0.5921 ms 67.2%
  triton_mm_11521 0.6190 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.0568 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0741 ms 100.0%
  triton_mm_8668 0.0755 ms 98.2%
  triton_mm_8662 0.0758 ms 97.7%
  triton_mm_8659 0.0761 ms 97.3%
  triton_mm_8661 0.0772 ms 96.0%
  triton_mm_8654 0.0778 ms 95.3%
  triton_mm_8665 0.0779 ms 95.1%
  triton_mm_8653 0.0779 ms 95.1%
  triton_mm_8666 0.0780 ms 95.0%
  mm 0.0783 ms 94.6%
SingleProcess AUTOTUNE benchmarking takes 1.8854 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0317 ms 100.0%
  triton_mm_8678 0.0319 ms 99.4%
  triton_mm_8680 0.0319 ms 99.4%
  triton_mm_8681 0.0319 ms 99.2%
  triton_mm_8676 0.0326 ms 97.2%
  triton_mm_8677 0.0327 ms 97.0%
  triton_mm_8673 0.0345 ms 91.9%
  triton_mm_8672 0.0349 ms 90.8%
  triton_mm_8671 0.0359 ms 88.2%
  triton_mm_8684 0.0361 ms 87.8%
SingleProcess AUTOTUNE benchmarking takes 2.1203 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9734 ms 100.0%
  triton_mm_8702 0.9798 ms 99.3%
  addmm 1.1033 ms 88.2%
  triton_mm_8699 1.1625 ms 83.7%
  triton_mm_8695 1.1849 ms 82.1%
  triton_mm_8703 1.2298 ms 79.2%
  triton_mm_8700 1.2806 ms 76.0%
  triton_mm_8696 1.2809 ms 76.0%
  triton_mm_8697 1.2950 ms 75.2%
  triton_mm_8692 1.5858 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.2973 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3301 ms 100.0%
  mm 0.3400 ms 97.1%
  triton_mm_8714 0.3960 ms 83.4%
  triton_mm_8718 0.4013 ms 82.3%
  triton_mm_8722 0.4037 ms 81.8%
  triton_mm_8715 0.4325 ms 76.3%
  triton_mm_8716 0.4333 ms 76.2%
  triton_mm_8719 0.4370 ms 75.5%
  triton_mm_8711 0.5253 ms 62.9%
  triton_mm_8712 0.5443 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0445 seconds and 0.0014 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2689 ms 100.0%
  mm 1.2714 ms 99.8%
  triton_mm_8777 1.4992 ms 84.6%
  triton_mm_8769 1.6052 ms 79.1%
  triton_mm_8773 1.6363 ms 77.6%
  triton_mm_8770 1.6822 ms 75.4%
  triton_mm_8771 1.6845 ms 75.3%
  triton_mm_8774 1.7108 ms 74.2%
  triton_mm_8766 2.1339 ms 59.5%
  triton_mm_8767 2.1392 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2250 seconds and 0.0016 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9186 0.3291 ms 100.0%
  mm 0.3397 ms 96.9%
  triton_mm_9179 0.3956 ms 83.2%
  triton_mm_9183 0.4011 ms 82.1%
  triton_mm_9187 0.4044 ms 81.4%
  triton_mm_9180 0.4315 ms 76.3%
  triton_mm_9181 0.4334 ms 75.9%
  triton_mm_9184 0.4365 ms 75.4%
  triton_mm_9176 0.5254 ms 62.6%
  triton_mm_9177 0.5441 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0441 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9241 1.2684 ms 100.0%
  mm 1.2708 ms 99.8%
  triton_mm_9242 1.4991 ms 84.6%
  triton_mm_9234 1.6172 ms 78.4%
  triton_mm_9238 1.6360 ms 77.5%
  triton_mm_9236 1.6827 ms 75.4%
  triton_mm_9239 1.6940 ms 74.9%
  triton_mm_9235 1.6963 ms 74.8%
  triton_mm_9231 2.1212 ms 59.8%
  triton_mm_9232 2.1382 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2208 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_9260 0.9733 ms 100.0%
  bias_addmm 0.9758 ms 99.7%
  addmm 1.1039 ms 88.2%
  triton_mm_9257 1.1612 ms 83.8%
  triton_mm_9253 1.1746 ms 82.9%
  triton_mm_9261 1.2205 ms 79.7%
  triton_mm_9258 1.2785 ms 76.1%
  triton_mm_9254 1.2831 ms 75.9%
  triton_mm_9255 1.2969 ms 75.0%
  triton_mm_9250 1.5854 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.2896 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0381 ms 100.0%
  triton_mm_11424 0.0390 ms 97.7%
  triton_mm_11428 0.0396 ms 96.2%
  triton_mm_11432 0.0399 ms 95.5%
  triton_mm_11437 0.0400 ms 95.3%
  triton_mm_11435 0.0407 ms 93.6%
  triton_mm_11422 0.0410 ms 93.0%
  triton_mm_11425 0.0413 ms 92.2%
  triton_mm_11431 0.0433 ms 88.0%
  mm 0.0449 ms 84.7%
SingleProcess AUTOTUNE benchmarking takes 1.8651 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3236 ms 100.0%
  bias_addmm 5.3377 ms 99.7%
  addmm 6.0046 ms 88.7%
  triton_mm_11489 6.5925 ms 80.8%
  triton_mm_11485 6.5992 ms 80.7%
  triton_mm_11493 6.6662 ms 79.9%
  triton_mm_11486 7.1886 ms 74.1%
  triton_mm_11490 7.1912 ms 74.0%
  triton_mm_11487 7.2686 ms 73.2%
  triton_mm_11482 8.9515 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.9569 seconds and 0.0012 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1750 ms 100.0%
  bias_addmm 0.1766 ms 99.1%
  addmm 0.1814 ms 96.5%
  triton_mm_11543 0.1933 ms 90.5%
  triton_mm_11544 0.1938 ms 90.3%
  triton_mm_11547 0.1939 ms 90.2%
  triton_mm_11542 0.1941 ms 90.2%
  triton_mm_11546 0.1947 ms 89.9%
  triton_mm_11550 0.2138 ms 81.8%
  triton_mm_11540 0.2352 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 2.1155 seconds and 0.0011 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [03:36<2:35:17, 216.69s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.36s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.37s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.37s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.37s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.37s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.37s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:57, 91.37s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [03:40<1:03:59, 91.41s/it] /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:20, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.27s/it]  /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [03:41<34:21, 50.29s/it]  
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.88s/it]
Validation Progress:   9%|9         | 4/44 [03:42<20:35, 30.89s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:43<13:05, 20.14s/it]
Validation Progress:  11%|#1        | 5/44 [03:44<13:05, 20.15s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  14%|#3        | 6/44 [03:45<08:39, 13.68s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  16%|#5        | 7/44 [03:46<05:54,  9.59s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  18%|#8        | 8/44 [03:47<04:08,  6.91s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  20%|##        | 9/44 [03:48<02:59,  5.13s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  20%|##        | 9/44 [03:48<03:00,  5.16s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:13,  3.93s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  23%|##2       | 10/44 [03:49<02:14,  3.95s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.10s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:41,  3.09s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.10s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.10s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.10s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.11s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.10s/it]
Validation Progress:  25%|##5       | 11/44 [03:51<01:42,  3.11s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.53s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.53s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.53s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:20,  2.52s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.54s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.53s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.53s/it]
Validation Progress:  27%|##7       | 12/44 [03:52<01:21,  2.54s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.14s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  30%|##9       | 13/44 [03:53<01:06,  2.15s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:55,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  32%|###1      | 14/44 [03:54<00:56,  1.87s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  34%|###4      | 15/44 [03:56<00:48,  1.67s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  36%|###6      | 16/44 [03:57<00:43,  1.56s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.49s/it]
Validation Progress:  39%|###8      | 17/44 [03:58<00:40,  1.48s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.41s/it]
Validation Progress:  41%|####      | 18/44 [03:59<00:36,  1.42s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:34,  1.40s/it]
Validation Progress:  43%|####3     | 19/44 [04:01<00:35,  1.41s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [04:02<00:32,  1.34s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.29s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.30s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.29s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.30s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.30s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.29s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.29s/it]
Validation Progress:  48%|####7     | 21/44 [04:03<00:29,  1.29s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [04:04<00:27,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [04:06<00:27,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [04:07<00:26,  1.30s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [04:08<00:25,  1.34s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.31s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [04:10<00:23,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:22,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:22,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:22,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:22,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [04:11<00:22,  1.30s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [04:12<00:20,  1.29s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [04:14<00:20,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [04:15<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [04:16<00:17,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [04:18<00:16,  1.34s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [04:19<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [04:20<00:12,  1.27s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [04:21<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [04:23<00:10,  1.30s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [04:24<00:09,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [04:26<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [04:27<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [04:28<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [04:29<00:03,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [04:31<00:02,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [04:32<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [04:33<00:00,  6.22s/it]

0it [00:00, ?it/s]
10it [00:00, 74.57it/s]
26it [00:00, 98.38it/s]
42it [00:00, 105.82it/s]
94it [00:00, 211.55it/s]
248it [00:00, 533.97it/s]
513it [00:00, 978.44it/s]
1025it [00:00, 1810.56it/s]
1537it [00:01, 2604.25it/s]
2049it [00:01, 3229.97it/s]
2561it [00:01, 3710.45it/s]
3073it [00:01, 4061.64it/s]
3585it [00:01, 4313.23it/s]
4097it [00:01, 4506.96it/s]
4609it [00:01, 4637.44it/s]
5121it [00:01, 4740.70it/s]
5633it [00:01, 4778.88it/s]
6145it [00:02, 4829.49it/s]
6657it [00:02, 4865.60it/s]
7169it [00:02, 4894.31it/s]
7681it [00:02, 4925.53it/s]
8193it [00:02, 4945.36it/s]
8705it [00:02, 4947.28it/s]
9217it [00:02, 4953.77it/s]
9729it [00:02, 4958.29it/s]
10241it [00:02, 4963.52it/s]
10753it [00:02, 4986.02it/s]
11344it [00:03, 5258.01it/s]
11679it [00:03, 3775.03it/s]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.24s/it]
100%|##########| 1/1 [00:03<00:00,  3.24s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|          | 1/157 [00:01<03:11,  1.23s/it]
  2%|1         | 3/157 [00:01<00:56,  2.70it/s]
  3%|3         | 5/157 [00:01<00:32,  4.66it/s]
  4%|4         | 7/157 [00:01<00:22,  6.72it/s]
  6%|5         | 9/157 [00:01<00:16,  8.74it/s]
  7%|7         | 11/157 [00:01<00:13, 10.57it/s]
  8%|8         | 13/157 [00:01<00:11, 12.03it/s]
 10%|9         | 15/157 [00:02<00:10, 13.35it/s]
 11%|#         | 17/157 [00:02<00:09, 14.43it/s]
 12%|#2        | 19/157 [00:02<00:09, 15.24it/s]
 13%|#3        | 21/157 [00:02<00:08, 15.33it/s]
 15%|#4        | 23/157 [00:02<00:08, 15.89it/s]
 16%|#5        | 25/157 [00:02<00:08, 16.34it/s]
 17%|#7        | 27/157 [00:02<00:07, 16.65it/s]
 18%|#8        | 29/157 [00:02<00:07, 16.85it/s]
 20%|#9        | 31/157 [00:03<00:07, 16.98it/s]
 21%|##1       | 33/157 [00:03<00:07, 17.11it/s]
 22%|##2       | 35/157 [00:03<00:07, 17.18it/s]
 24%|##3       | 37/157 [00:03<00:07, 16.44it/s]
 25%|##4       | 39/157 [00:03<00:07, 16.66it/s]
 26%|##6       | 41/157 [00:03<00:06, 16.89it/s]
 27%|##7       | 43/157 [00:03<00:06, 17.06it/s]
 29%|##8       | 45/157 [00:03<00:06, 17.16it/s]
 30%|##9       | 47/157 [00:03<00:06, 17.22it/s]
 31%|###1      | 49/157 [00:04<00:06, 17.27it/s]
 32%|###2      | 51/157 [00:04<00:06, 17.31it/s]
 34%|###3      | 53/157 [00:04<00:05, 17.34it/s]
 35%|###5      | 55/157 [00:04<00:05, 17.34it/s]
 36%|###6      | 57/157 [00:04<00:05, 17.24it/s]
 38%|###7      | 59/157 [00:04<00:05, 17.16it/s]
 39%|###8      | 61/157 [00:04<00:05, 17.10it/s]
 40%|####      | 63/157 [00:04<00:05, 17.09it/s]
 41%|####1     | 65/157 [00:05<00:05, 17.03it/s]
 43%|####2     | 67/157 [00:05<00:05, 16.79it/s]
 44%|####3     | 69/157 [00:05<00:05, 15.46it/s]
 45%|####5     | 71/157 [00:05<00:05, 15.88it/s]
 46%|####6     | 73/157 [00:05<00:05, 16.28it/s]
 48%|####7     | 75/157 [00:05<00:04, 16.57it/s]
 49%|####9     | 77/157 [00:05<00:04, 16.76it/s]
 50%|#####     | 79/157 [00:05<00:04, 16.91it/s]
 52%|#####1    | 81/157 [00:05<00:04, 17.01it/s]
 53%|#####2    | 83/157 [00:06<00:04, 17.10it/s]
 54%|#####4    | 85/157 [00:06<00:04, 17.27it/s]
 55%|#####5    | 87/157 [00:06<00:04, 17.17it/s]
 57%|#####6    | 89/157 [00:06<00:03, 17.21it/s]
 58%|#####7    | 91/157 [00:06<00:03, 17.37it/s]
 59%|#####9    | 93/157 [00:06<00:03, 17.23it/s]
 61%|######    | 95/157 [00:06<00:03, 17.21it/s]
 62%|######1   | 97/157 [00:06<00:03, 17.18it/s]
 63%|######3   | 99/157 [00:07<00:03, 17.02it/s]
 64%|######4   | 101/157 [00:07<00:03, 16.86it/s]
 66%|######5   | 103/157 [00:07<00:03, 16.85it/s]
 67%|######6   | 105/157 [00:07<00:03, 16.80it/s]
 68%|######8   | 107/157 [00:07<00:02, 17.08it/s]
 69%|######9   | 109/157 [00:07<00:02, 17.12it/s]
 71%|#######   | 111/157 [00:07<00:02, 16.93it/s]
 72%|#######1  | 113/157 [00:07<00:02, 16.83it/s]
 73%|#######3  | 115/157 [00:07<00:02, 16.82it/s]
 75%|#######4  | 117/157 [00:08<00:02, 16.91it/s]
 76%|#######5  | 119/157 [00:08<00:02, 16.94it/s]
 77%|#######7  | 121/157 [00:08<00:02, 17.01it/s]
 78%|#######8  | 123/157 [00:08<00:01, 17.06it/s]
 80%|#######9  | 125/157 [00:08<00:01, 17.21it/s]
 81%|########  | 127/157 [00:08<00:01, 17.03it/s]
 82%|########2 | 129/157 [00:08<00:01, 17.08it/s]
 83%|########3 | 131/157 [00:08<00:01, 17.13it/s]
 85%|########4 | 133/157 [00:09<00:01, 14.74it/s]
 86%|########5 | 135/157 [00:09<00:01, 15.26it/s]
 87%|########7 | 137/157 [00:09<00:01, 15.79it/s]
 89%|########8 | 139/157 [00:09<00:01, 16.32it/s]
 90%|########9 | 141/157 [00:09<00:00, 16.44it/s]
 91%|#########1| 143/157 [00:09<00:00, 16.66it/s]
 92%|#########2| 145/157 [00:09<00:00, 16.80it/s]
 94%|#########3| 147/157 [00:09<00:00, 16.94it/s]
 95%|#########4| 149/157 [00:10<00:00, 17.01it/s]
 96%|#########6| 151/157 [00:10<00:00, 17.21it/s]
 97%|#########7| 153/157 [00:10<00:00, 17.05it/s]
 99%|#########8| 155/157 [00:10<00:00, 17.12it/s]
100%|##########| 157/157 [00:11<00:00,  6.01it/s]
100%|##########| 157/157 [00:11<00:00, 14.00it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:45, 17.27it/s]
  1%|          | 4/781 [00:00<00:44, 17.33it/s]
  1%|          | 6/781 [00:00<00:45, 16.94it/s]
  1%|1         | 8/781 [00:00<00:46, 16.62it/s]
  1%|1         | 10/781 [00:00<00:47, 16.40it/s]
  2%|1         | 12/781 [00:00<00:47, 16.34it/s]
  2%|1         | 14/781 [00:00<00:46, 16.55it/s]
  2%|2         | 16/781 [00:00<00:46, 16.42it/s]
  2%|2         | 18/781 [00:01<00:46, 16.34it/s]
  3%|2         | 20/781 [00:01<00:46, 16.43it/s]
  3%|2         | 22/781 [00:01<00:46, 16.34it/s]
  3%|3         | 24/781 [00:01<00:46, 16.41it/s]
  3%|3         | 26/781 [00:01<00:45, 16.47it/s]
  4%|3         | 28/781 [00:01<00:46, 16.37it/s]
  4%|3         | 30/781 [00:01<00:45, 16.54it/s]
  4%|4         | 32/781 [00:01<00:45, 16.56it/s]
  4%|4         | 34/781 [00:02<00:44, 16.65it/s]
  5%|4         | 36/781 [00:02<00:45, 16.51it/s]
  5%|4         | 38/781 [00:02<00:45, 16.43it/s]
  5%|5         | 40/781 [00:02<00:44, 16.47it/s]
  5%|5         | 42/781 [00:02<00:44, 16.61it/s]
  6%|5         | 44/781 [00:02<00:44, 16.59it/s]
  6%|5         | 46/781 [00:02<00:44, 16.46it/s]
  6%|6         | 48/781 [00:02<00:44, 16.35it/s]
  6%|6         | 50/781 [00:03<00:44, 16.45it/s]
  7%|6         | 52/781 [00:03<00:44, 16.48it/s]
  7%|6         | 54/781 [00:03<00:43, 16.59it/s]
  7%|7         | 56/781 [00:03<00:43, 16.56it/s]
  7%|7         | 58/781 [00:03<00:43, 16.67it/s]
  8%|7         | 60/781 [00:03<00:43, 16.47it/s]
  8%|7         | 62/781 [00:03<00:43, 16.50it/s]
  8%|8         | 64/781 [00:03<00:43, 16.51it/s]
  8%|8         | 66/781 [00:03<00:43, 16.53it/s]
  9%|8         | 68/781 [00:04<00:43, 16.52it/s]
  9%|8         | 70/781 [00:04<00:43, 16.43it/s]
  9%|9         | 72/781 [00:04<00:43, 16.48it/s]
  9%|9         | 74/781 [00:04<00:42, 16.50it/s]
 10%|9         | 76/781 [00:04<00:43, 16.23it/s]
 10%|9         | 78/781 [00:04<00:43, 16.32it/s]
 10%|#         | 80/781 [00:04<00:42, 16.37it/s]
 10%|#         | 82/781 [00:04<00:42, 16.54it/s]
 11%|#         | 84/781 [00:05<00:42, 16.55it/s]
 11%|#1        | 86/781 [00:05<00:41, 16.59it/s]
 11%|#1        | 88/781 [00:05<00:42, 16.45it/s]
 12%|#1        | 90/781 [00:05<00:41, 16.49it/s]
 12%|#1        | 92/781 [00:05<00:42, 16.33it/s]
 12%|#2        | 94/781 [00:05<00:42, 16.28it/s]
 12%|#2        | 96/781 [00:05<00:41, 16.33it/s]
 13%|#2        | 98/781 [00:05<00:41, 16.38it/s]
 13%|#2        | 100/781 [00:06<00:41, 16.38it/s]
 13%|#3        | 102/781 [00:06<00:41, 16.42it/s]
 13%|#3        | 104/781 [00:06<00:41, 16.37it/s]
 14%|#3        | 106/781 [00:06<00:41, 16.40it/s]
 14%|#3        | 108/781 [00:06<00:41, 16.29it/s]
 14%|#4        | 110/781 [00:06<00:41, 16.29it/s]
 14%|#4        | 112/781 [00:06<00:41, 16.24it/s]
 15%|#4        | 114/781 [00:06<00:40, 16.33it/s]
 15%|#4        | 116/781 [00:07<00:40, 16.40it/s]
 15%|#5        | 118/781 [00:07<00:40, 16.55it/s]
 15%|#5        | 120/781 [00:07<00:39, 16.57it/s]
 16%|#5        | 122/781 [00:07<00:40, 16.47it/s]
 16%|#5        | 124/781 [00:07<00:39, 16.45it/s]
 16%|#6        | 126/781 [00:07<00:40, 16.35it/s]
 16%|#6        | 128/781 [00:07<00:39, 16.33it/s]
 17%|#6        | 130/781 [00:07<00:40, 16.27it/s]
 17%|#6        | 132/781 [00:08<00:39, 16.24it/s]
 17%|#7        | 134/781 [00:08<00:39, 16.19it/s]
 17%|#7        | 136/781 [00:08<00:39, 16.25it/s]
 18%|#7        | 138/781 [00:08<00:39, 16.35it/s]
 18%|#7        | 140/781 [00:08<00:39, 16.40it/s]
 18%|#8        | 142/781 [00:08<00:38, 16.42it/s]
 18%|#8        | 144/781 [00:08<00:38, 16.55it/s]
 19%|#8        | 146/781 [00:08<00:38, 16.42it/s]
 19%|#8        | 148/781 [00:09<00:38, 16.32it/s]
 19%|#9        | 150/781 [00:09<00:38, 16.39it/s]
 19%|#9        | 152/781 [00:09<00:38, 16.37it/s]
 20%|#9        | 154/781 [00:09<00:38, 16.30it/s]
 20%|#9        | 156/781 [00:09<00:37, 16.46it/s]
 20%|##        | 158/781 [00:09<00:37, 16.46it/s]
 20%|##        | 160/781 [00:09<00:37, 16.59it/s]
 21%|##        | 162/781 [00:09<00:37, 16.67it/s]
 21%|##        | 164/781 [00:09<00:37, 16.63it/s]
 21%|##1       | 166/781 [00:10<00:36, 16.66it/s]
 22%|##1       | 168/781 [00:10<00:36, 16.66it/s]
 22%|##1       | 170/781 [00:10<00:36, 16.74it/s]
 22%|##2       | 172/781 [00:10<00:36, 16.76it/s]
 22%|##2       | 174/781 [00:10<00:36, 16.66it/s]
 23%|##2       | 176/781 [00:10<00:36, 16.60it/s]
 23%|##2       | 178/781 [00:10<00:36, 16.69it/s]
 23%|##3       | 180/781 [00:10<00:35, 16.74it/s]
 23%|##3       | 182/781 [00:11<00:35, 16.68it/s]
 24%|##3       | 184/781 [00:11<00:36, 16.50it/s]
 24%|##3       | 186/781 [00:11<00:35, 16.63it/s]
 24%|##4       | 188/781 [00:11<00:35, 16.48it/s]
 24%|##4       | 190/781 [00:11<00:36, 16.40it/s]
 25%|##4       | 192/781 [00:11<00:35, 16.44it/s]
 25%|##4       | 194/781 [00:11<00:36, 16.28it/s]
 25%|##5       | 196/781 [00:11<00:35, 16.34it/s]
 25%|##5       | 198/781 [00:12<00:35, 16.28it/s]
 26%|##5       | 200/781 [00:12<00:35, 16.38it/s]
 26%|##5       | 202/781 [00:12<00:35, 16.44it/s]
 26%|##6       | 204/781 [00:12<00:35, 16.47it/s]
 26%|##6       | 206/781 [00:12<00:34, 16.53it/s]
 27%|##6       | 208/781 [00:12<00:34, 16.62it/s]
 27%|##6       | 210/781 [00:12<00:34, 16.60it/s]
 27%|##7       | 212/781 [00:12<00:34, 16.73it/s]
 27%|##7       | 214/781 [00:12<00:33, 16.68it/s]
 28%|##7       | 216/781 [00:13<00:33, 16.65it/s]
 28%|##7       | 218/781 [00:13<00:33, 16.76it/s]
 28%|##8       | 220/781 [00:13<00:33, 16.78it/s]
 28%|##8       | 222/781 [00:13<00:33, 16.69it/s]
 29%|##8       | 224/781 [00:13<00:33, 16.67it/s]
 29%|##8       | 226/781 [00:13<00:33, 16.75it/s]
 29%|##9       | 228/781 [00:13<00:33, 16.67it/s]
 29%|##9       | 230/781 [00:13<00:33, 16.53it/s]
 30%|##9       | 232/781 [00:14<00:33, 16.43it/s]
 30%|##9       | 234/781 [00:14<00:33, 16.35it/s]
 30%|###       | 236/781 [00:14<00:33, 16.32it/s]
 30%|###       | 238/781 [00:14<00:33, 16.42it/s]
 31%|###       | 240/781 [00:14<00:32, 16.60it/s]
 31%|###       | 242/781 [00:14<00:32, 16.60it/s]
 31%|###1      | 244/781 [00:14<00:32, 16.71it/s]
 31%|###1      | 246/781 [00:14<00:32, 16.66it/s]
 32%|###1      | 248/781 [00:15<00:32, 16.56it/s]
 32%|###2      | 250/781 [00:15<00:32, 16.48it/s]
 32%|###2      | 252/781 [00:15<00:32, 16.51it/s]
 33%|###2      | 254/781 [00:15<00:31, 16.56it/s]
 33%|###2      | 256/781 [00:15<00:31, 16.64it/s]
 33%|###3      | 258/781 [00:15<00:31, 16.57it/s]
 33%|###3      | 260/781 [00:15<00:31, 16.36it/s]
 34%|###3      | 262/781 [00:16<00:42, 12.26it/s]
 34%|###3      | 264/781 [00:16<00:39, 13.16it/s]
 34%|###4      | 266/781 [00:16<00:36, 13.94it/s]
 34%|###4      | 268/781 [00:16<00:35, 14.51it/s]
 35%|###4      | 270/781 [00:16<00:33, 15.12it/s]
 35%|###4      | 272/781 [00:16<00:32, 15.58it/s]
 35%|###5      | 274/781 [00:16<00:32, 15.77it/s]
 35%|###5      | 276/781 [00:16<00:31, 15.92it/s]
 36%|###5      | 278/781 [00:17<00:31, 16.17it/s]
 36%|###5      | 280/781 [00:17<00:30, 16.29it/s]
 36%|###6      | 282/781 [00:17<00:30, 16.36it/s]
 36%|###6      | 284/781 [00:17<00:30, 16.33it/s]
 37%|###6      | 286/781 [00:17<00:30, 16.29it/s]
 37%|###6      | 288/781 [00:17<00:30, 16.28it/s]
 37%|###7      | 290/781 [00:17<00:30, 16.27it/s]
 37%|###7      | 292/781 [00:17<00:30, 16.27it/s]
 38%|###7      | 294/781 [00:17<00:29, 16.46it/s]
 38%|###7      | 296/781 [00:18<00:29, 16.47it/s]
 38%|###8      | 298/781 [00:18<00:29, 16.40it/s]
 38%|###8      | 300/781 [00:18<00:29, 16.40it/s]
 39%|###8      | 302/781 [00:18<00:29, 16.19it/s]
 39%|###8      | 304/781 [00:18<00:29, 16.16it/s]
 39%|###9      | 306/781 [00:18<00:29, 16.19it/s]
 39%|###9      | 308/781 [00:18<00:29, 16.20it/s]
 40%|###9      | 310/781 [00:18<00:29, 16.22it/s]
 40%|###9      | 312/781 [00:19<00:28, 16.23it/s]
 40%|####      | 314/781 [00:19<00:28, 16.26it/s]
 40%|####      | 316/781 [00:19<00:28, 16.22it/s]
 41%|####      | 318/781 [00:19<00:28, 16.12it/s]
 41%|####      | 320/781 [00:19<00:28, 16.27it/s]
 41%|####1     | 322/781 [00:19<00:28, 16.39it/s]
 41%|####1     | 324/781 [00:19<00:27, 16.35it/s]
 42%|####1     | 326/781 [00:19<00:27, 16.32it/s]
 42%|####1     | 328/781 [00:20<00:27, 16.30it/s]
 42%|####2     | 330/781 [00:20<00:27, 16.28it/s]
 43%|####2     | 332/781 [00:20<00:27, 16.30it/s]
 43%|####2     | 334/781 [00:20<00:27, 16.28it/s]
 43%|####3     | 336/781 [00:20<00:27, 16.40it/s]
 43%|####3     | 338/781 [00:20<00:27, 16.35it/s]
 44%|####3     | 340/781 [00:20<00:27, 16.33it/s]
 44%|####3     | 342/781 [00:20<00:26, 16.31it/s]
 44%|####4     | 344/781 [00:21<00:26, 16.33it/s]
 44%|####4     | 346/781 [00:21<00:26, 16.28it/s]
 45%|####4     | 348/781 [00:21<00:26, 16.17it/s]
 45%|####4     | 350/781 [00:21<00:26, 16.10it/s]
 45%|####5     | 352/781 [00:21<00:26, 16.05it/s]
 45%|####5     | 354/781 [00:21<00:26, 16.08it/s]
 46%|####5     | 356/781 [00:21<00:26, 15.99it/s]
 46%|####5     | 358/781 [00:21<00:26, 15.95it/s]
 46%|####6     | 360/781 [00:22<00:26, 15.92it/s]
 46%|####6     | 362/781 [00:22<00:26, 15.94it/s]
 47%|####6     | 364/781 [00:22<00:25, 16.06it/s]
 47%|####6     | 366/781 [00:22<00:25, 16.22it/s]
 47%|####7     | 368/781 [00:22<00:25, 16.43it/s]
 47%|####7     | 370/781 [00:22<00:25, 16.37it/s]
 48%|####7     | 372/781 [00:22<00:25, 16.33it/s]
 48%|####7     | 374/781 [00:22<00:24, 16.34it/s]
 48%|####8     | 376/781 [00:23<00:24, 16.29it/s]
 48%|####8     | 378/781 [00:23<00:24, 16.32it/s]
 49%|####8     | 380/781 [00:23<00:24, 16.30it/s]
 49%|####8     | 382/781 [00:23<00:24, 16.28it/s]
 49%|####9     | 384/781 [00:23<00:24, 16.42it/s]
 49%|####9     | 386/781 [00:23<00:23, 16.47it/s]
 50%|####9     | 388/781 [00:23<00:23, 16.41it/s]
 50%|####9     | 390/781 [00:23<00:23, 16.41it/s]
 50%|#####     | 392/781 [00:24<00:23, 16.48it/s]
 50%|#####     | 394/781 [00:24<00:23, 16.43it/s]
 51%|#####     | 396/781 [00:24<00:23, 16.37it/s]
 51%|#####     | 398/781 [00:24<00:23, 16.44it/s]
 51%|#####1    | 400/781 [00:24<00:23, 16.40it/s]
 51%|#####1    | 402/781 [00:24<00:23, 16.40it/s]
 52%|#####1    | 404/781 [00:24<00:22, 16.44it/s]
 52%|#####1    | 406/781 [00:24<00:22, 16.49it/s]
 52%|#####2    | 408/781 [00:24<00:22, 16.45it/s]
 52%|#####2    | 410/781 [00:25<00:22, 16.43it/s]
 53%|#####2    | 412/781 [00:25<00:22, 16.41it/s]
 53%|#####3    | 414/781 [00:25<00:22, 16.37it/s]
 53%|#####3    | 416/781 [00:25<00:22, 16.53it/s]
 54%|#####3    | 418/781 [00:25<00:22, 16.46it/s]
 54%|#####3    | 420/781 [00:25<00:22, 16.36it/s]
 54%|#####4    | 422/781 [00:25<00:21, 16.56it/s]
 54%|#####4    | 424/781 [00:25<00:21, 16.47it/s]
 55%|#####4    | 426/781 [00:26<00:21, 16.62it/s]
 55%|#####4    | 428/781 [00:26<00:21, 16.60it/s]
 55%|#####5    | 430/781 [00:26<00:21, 16.52it/s]
 55%|#####5    | 432/781 [00:26<00:21, 16.37it/s]
 56%|#####5    | 434/781 [00:26<00:21, 16.31it/s]
 56%|#####5    | 436/781 [00:26<00:21, 16.22it/s]
 56%|#####6    | 438/781 [00:26<00:20, 16.34it/s]
 56%|#####6    | 440/781 [00:26<00:20, 16.39it/s]
 57%|#####6    | 442/781 [00:27<00:20, 16.48it/s]
 57%|#####6    | 444/781 [00:27<00:20, 16.56it/s]
 57%|#####7    | 446/781 [00:27<00:20, 16.61it/s]
 57%|#####7    | 448/781 [00:27<00:19, 16.67it/s]
 58%|#####7    | 450/781 [00:27<00:19, 16.66it/s]
 58%|#####7    | 452/781 [00:27<00:19, 16.68it/s]
 58%|#####8    | 454/781 [00:27<00:19, 16.58it/s]
 58%|#####8    | 456/781 [00:27<00:19, 16.62it/s]
 59%|#####8    | 458/781 [00:28<00:19, 16.66it/s]
 59%|#####8    | 460/781 [00:28<00:19, 16.62it/s]
 59%|#####9    | 462/781 [00:28<00:19, 16.55it/s]
 59%|#####9    | 464/781 [00:28<00:18, 16.74it/s]
 60%|#####9    | 466/781 [00:28<00:18, 16.84it/s]
 60%|#####9    | 468/781 [00:28<00:18, 16.81it/s]
 60%|######    | 470/781 [00:28<00:18, 16.82it/s]
 60%|######    | 472/781 [00:29<00:29, 10.63it/s]
 61%|######    | 474/781 [00:29<00:26, 11.71it/s]
 61%|######    | 476/781 [00:29<00:24, 12.62it/s]
 61%|######1   | 478/781 [00:29<00:22, 13.45it/s]
 61%|######1   | 480/781 [00:29<00:21, 14.09it/s]
 62%|######1   | 482/781 [00:29<00:20, 14.45it/s]
 62%|######1   | 484/781 [00:29<00:20, 14.76it/s]
 62%|######2   | 486/781 [00:29<00:19, 14.93it/s]
 62%|######2   | 488/781 [00:30<00:19, 15.06it/s]
 63%|######2   | 490/781 [00:30<00:18, 15.33it/s]
 63%|######2   | 492/781 [00:30<00:18, 15.56it/s]
 63%|######3   | 494/781 [00:30<00:18, 15.72it/s]
 64%|######3   | 496/781 [00:30<00:18, 15.61it/s]
 64%|######3   | 498/781 [00:30<00:17, 15.73it/s]
 64%|######4   | 500/781 [00:30<00:17, 15.87it/s]
 64%|######4   | 502/781 [00:30<00:17, 15.64it/s]
 65%|######4   | 504/781 [00:31<00:17, 15.46it/s]
 65%|######4   | 506/781 [00:31<00:17, 15.33it/s]
 65%|######5   | 508/781 [00:31<00:17, 15.44it/s]
 65%|######5   | 510/781 [00:31<00:17, 15.51it/s]
 66%|######5   | 512/781 [00:31<00:17, 15.32it/s]
 66%|######5   | 514/781 [00:31<00:17, 15.66it/s]
 66%|######6   | 516/781 [00:31<00:16, 15.97it/s]
 66%|######6   | 518/781 [00:32<00:16, 15.77it/s]
 67%|######6   | 520/781 [00:32<00:27,  9.59it/s]
 67%|######6   | 522/781 [00:32<00:23, 10.90it/s]
 67%|######7   | 524/781 [00:32<00:21, 11.94it/s]
 67%|######7   | 526/781 [00:32<00:19, 12.80it/s]
 68%|######7   | 528/781 [00:32<00:18, 13.46it/s]
 68%|######7   | 530/781 [00:33<00:17, 14.14it/s]
 68%|######8   | 532/781 [00:33<00:16, 14.68it/s]
 68%|######8   | 534/781 [00:33<00:16, 14.93it/s]
 69%|######8   | 536/781 [00:33<00:16, 15.26it/s]
 69%|######8   | 538/781 [00:33<00:15, 15.58it/s]
 69%|######9   | 540/781 [00:33<00:15, 15.56it/s]
 69%|######9   | 542/781 [00:33<00:14, 15.94it/s]
 70%|######9   | 544/781 [00:33<00:14, 16.02it/s]
 70%|######9   | 546/781 [00:34<00:14, 15.85it/s]
 70%|#######   | 548/781 [00:34<00:14, 15.73it/s]
 70%|#######   | 550/781 [00:34<00:14, 15.63it/s]
 71%|#######   | 552/781 [00:34<00:14, 15.56it/s]
 71%|#######   | 554/781 [00:34<00:14, 15.69it/s]
 71%|#######1  | 556/781 [00:34<00:14, 15.58it/s]
 71%|#######1  | 558/781 [00:34<00:14, 15.69it/s]
 72%|#######1  | 560/781 [00:34<00:13, 15.83it/s]
 72%|#######1  | 562/781 [00:35<00:13, 15.88it/s]
 72%|#######2  | 564/781 [00:35<00:13, 15.64it/s]
 72%|#######2  | 566/781 [00:35<00:14, 15.35it/s]
 73%|#######2  | 568/781 [00:35<00:13, 15.29it/s]
 73%|#######2  | 570/781 [00:35<00:13, 15.30it/s]
 73%|#######3  | 572/781 [00:35<00:13, 15.41it/s]
 73%|#######3  | 574/781 [00:35<00:13, 15.66it/s]
 74%|#######3  | 576/781 [00:35<00:12, 15.81it/s]
 74%|#######4  | 578/781 [00:36<00:12, 16.05it/s]
 74%|#######4  | 580/781 [00:36<00:12, 15.83it/s]
 75%|#######4  | 582/781 [00:36<00:12, 15.85it/s]
 75%|#######4  | 584/781 [00:36<00:12, 15.92it/s]
 75%|#######5  | 586/781 [00:36<00:12, 15.84it/s]
 75%|#######5  | 588/781 [00:36<00:12, 15.73it/s]
 76%|#######5  | 590/781 [00:36<00:12, 15.66it/s]
 76%|#######5  | 592/781 [00:36<00:12, 15.66it/s]
 76%|#######6  | 594/781 [00:37<00:11, 15.69it/s]
 76%|#######6  | 596/781 [00:37<00:11, 15.81it/s]
 77%|#######6  | 598/781 [00:37<00:11, 15.91it/s]
 77%|#######6  | 600/781 [00:37<00:11, 15.96it/s]
 77%|#######7  | 602/781 [00:37<00:11, 16.22it/s]
 77%|#######7  | 604/781 [00:37<00:10, 16.40it/s]
 78%|#######7  | 606/781 [00:37<00:10, 16.29it/s]
 78%|#######7  | 608/781 [00:37<00:10, 16.20it/s]
 78%|#######8  | 610/781 [00:38<00:10, 16.18it/s]
 78%|#######8  | 612/781 [00:38<00:10, 15.94it/s]
 79%|#######8  | 614/781 [00:38<00:10, 15.78it/s]
 79%|#######8  | 616/781 [00:38<00:10, 16.09it/s]
 79%|#######9  | 618/781 [00:38<00:10, 16.08it/s]
 79%|#######9  | 620/781 [00:38<00:10, 16.07it/s]
 80%|#######9  | 622/781 [00:38<00:09, 16.09it/s]
 80%|#######9  | 624/781 [00:38<00:09, 16.10it/s]
 80%|########  | 626/781 [00:39<00:09, 16.31it/s]
 80%|########  | 628/781 [00:39<00:09, 16.27it/s]
 81%|########  | 630/781 [00:39<00:09, 16.23it/s]
 81%|########  | 632/781 [00:39<00:09, 16.01it/s]
 81%|########1 | 634/781 [00:39<00:09, 15.84it/s]
 81%|########1 | 636/781 [00:39<00:09, 15.71it/s]
 82%|########1 | 638/781 [00:39<00:09, 15.83it/s]
 82%|########1 | 640/781 [00:39<00:08, 15.93it/s]
 82%|########2 | 642/781 [00:40<00:08, 16.17it/s]
 82%|########2 | 644/781 [00:40<00:08, 16.39it/s]
 83%|########2 | 646/781 [00:40<00:08, 16.05it/s]
 83%|########2 | 648/781 [00:40<00:08, 16.01it/s]
 83%|########3 | 650/781 [00:40<00:08, 15.75it/s]
 83%|########3 | 652/781 [00:40<00:08, 15.84it/s]
 84%|########3 | 654/781 [00:40<00:07, 15.93it/s]
 84%|########3 | 656/781 [00:40<00:07, 16.03it/s]
 84%|########4 | 658/781 [00:41<00:07, 16.09it/s]
 85%|########4 | 660/781 [00:41<00:07, 15.92it/s]
 85%|########4 | 662/781 [00:41<00:07, 15.80it/s]
 85%|########5 | 664/781 [00:41<00:07, 15.85it/s]
 85%|########5 | 666/781 [00:41<00:07, 15.74it/s]
 86%|########5 | 668/781 [00:41<00:07, 16.02it/s]
 86%|########5 | 670/781 [00:41<00:06, 16.05it/s]
 86%|########6 | 672/781 [00:41<00:06, 16.05it/s]
 86%|########6 | 674/781 [00:42<00:06, 16.05it/s]
 87%|########6 | 676/781 [00:42<00:06, 15.88it/s]
 87%|########6 | 678/781 [00:42<00:06, 15.95it/s]
 87%|########7 | 680/781 [00:42<00:06, 16.01it/s]
 87%|########7 | 682/781 [00:42<00:06, 15.88it/s]
 88%|########7 | 684/781 [00:42<00:06, 15.78it/s]
 88%|########7 | 686/781 [00:42<00:05, 15.91it/s]
 88%|########8 | 688/781 [00:42<00:05, 16.22it/s]
 88%|########8 | 690/781 [00:43<00:05, 16.22it/s]
 89%|########8 | 692/781 [00:43<00:05, 16.01it/s]
 89%|########8 | 694/781 [00:43<00:05, 16.08it/s]
 89%|########9 | 696/781 [00:43<00:05, 16.14it/s]
 89%|########9 | 698/781 [00:43<00:05, 16.22it/s]
 90%|########9 | 700/781 [00:43<00:04, 16.44it/s]
 90%|########9 | 702/781 [00:43<00:04, 16.39it/s]
 90%|######### | 704/781 [00:43<00:04, 16.16it/s]
 90%|######### | 706/781 [00:44<00:04, 16.02it/s]
 91%|######### | 708/781 [00:44<00:04, 15.90it/s]
 91%|######### | 710/781 [00:44<00:04, 15.81it/s]
 91%|#########1| 712/781 [00:44<00:04, 15.77it/s]
 91%|#########1| 714/781 [00:44<00:04, 15.74it/s]
 92%|#########1| 716/781 [00:44<00:04, 15.73it/s]
 92%|#########1| 718/781 [00:44<00:04, 15.71it/s]
 92%|#########2| 720/781 [00:44<00:03, 15.70it/s]
 92%|#########2| 722/781 [00:45<00:03, 15.70it/s]
 93%|#########2| 724/781 [00:45<00:03, 15.71it/s]
 93%|#########2| 726/781 [00:45<00:03, 15.68it/s]
 93%|#########3| 728/781 [00:45<00:03, 15.55it/s]
 93%|#########3| 730/781 [00:45<00:03, 15.45it/s]
 94%|#########3| 732/781 [00:45<00:03, 15.12it/s]
 94%|#########3| 734/781 [00:45<00:03, 15.14it/s]
 94%|#########4| 736/781 [00:46<00:02, 15.22it/s]
 94%|#########4| 738/781 [00:46<00:02, 15.16it/s]
 95%|#########4| 740/781 [00:46<00:02, 15.17it/s]
 95%|#########5| 742/781 [00:46<00:02, 15.27it/s]
 95%|#########5| 744/781 [00:46<00:02, 15.28it/s]
 96%|#########5| 746/781 [00:46<00:02, 15.31it/s]
 96%|#########5| 748/781 [00:46<00:02, 15.39it/s]
 96%|#########6| 750/781 [00:46<00:02, 15.46it/s]
 96%|#########6| 752/781 [00:47<00:01, 15.52it/s]
 97%|#########6| 754/781 [00:47<00:01, 15.55it/s]
 97%|#########6| 756/781 [00:47<00:01, 15.56it/s]
 97%|#########7| 758/781 [00:47<00:01, 15.55it/s]
 97%|#########7| 760/781 [00:47<00:01, 15.56it/s]
 98%|#########7| 762/781 [00:47<00:01, 15.56it/s]
 98%|#########7| 764/781 [00:47<00:01, 15.58it/s]
 98%|#########8| 766/781 [00:47<00:00, 15.58it/s]
 98%|#########8| 768/781 [00:48<00:00, 15.61it/s]
 99%|#########8| 770/781 [00:48<00:00, 15.58it/s]
 99%|#########8| 772/781 [00:48<00:00, 15.58it/s]
 99%|#########9| 774/781 [00:48<00:00, 15.60it/s]
 99%|#########9| 776/781 [00:48<00:00, 15.62it/s]
100%|#########9| 778/781 [00:48<00:00, 15.54it/s]
100%|#########9| 780/781 [00:48<00:00, 15.46it/s]
100%|##########| 781/781 [00:48<00:00, 15.96it/s]
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.26s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.26s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.28s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.28s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:55,  1.28s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:55,  1.28s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:55,  1.29s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:56,  1.31s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.16s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.16s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.17s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.17s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.17s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.17s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.17s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:44,  1.15s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:44,  1.15s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.15s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.15s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.16s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.15s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.16s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:46,  1.18s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:41,  1.19s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.21s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.22s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:40,  1.21s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.22s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.22s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.21s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.25s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.26s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.26s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.43s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:59<00:00,  2.44s/it]
Validation Progress: 100%|##########| 44/44 [00:59<00:00,  1.36s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:02<00:00,  2.25s/it]
100%|##########| 1/1 [00:02<00:00,  2.25s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.82it/s]
  3%|2         | 4/157 [00:00<00:10, 15.00it/s]
  4%|3         | 6/157 [00:00<00:09, 15.72it/s]
  5%|5         | 8/157 [00:00<00:09, 16.10it/s]
  6%|6         | 10/157 [00:00<00:08, 16.36it/s]
  8%|7         | 12/157 [00:00<00:08, 16.43it/s]
  9%|8         | 14/157 [00:00<00:08, 16.64it/s]
 10%|#         | 16/157 [00:00<00:08, 16.71it/s]
 11%|#1        | 18/157 [00:01<00:08, 16.87it/s]
 13%|#2        | 20/157 [00:01<00:08, 16.84it/s]
 14%|#4        | 22/157 [00:01<00:07, 16.90it/s]
 15%|#5        | 24/157 [00:01<00:07, 16.98it/s]
 17%|#6        | 26/157 [00:01<00:07, 17.06it/s]
 18%|#7        | 28/157 [00:01<00:07, 16.95it/s]
 19%|#9        | 30/157 [00:01<00:07, 16.86it/s]
 20%|##        | 32/157 [00:01<00:07, 16.89it/s]
 22%|##1       | 34/157 [00:02<00:07, 16.96it/s]
 23%|##2       | 36/157 [00:02<00:07, 17.05it/s]
 24%|##4       | 38/157 [00:02<00:06, 17.02it/s]
 25%|##5       | 40/157 [00:02<00:06, 16.96it/s]
 27%|##6       | 42/157 [00:02<00:06, 17.02it/s]
 28%|##8       | 44/157 [00:02<00:06, 17.05it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.78it/s]
 31%|###       | 48/157 [00:02<00:06, 16.79it/s]
 32%|###1      | 50/157 [00:02<00:06, 16.82it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.78it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.68it/s]
 36%|###5      | 56/157 [00:03<00:06, 16.71it/s]
 37%|###6      | 58/157 [00:03<00:05, 16.81it/s]
 38%|###8      | 60/157 [00:03<00:05, 16.76it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.86it/s]
 41%|####      | 64/157 [00:03<00:05, 16.96it/s]
 42%|####2     | 66/157 [00:03<00:05, 16.98it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.96it/s]
 45%|####4     | 70/157 [00:04<00:05, 17.07it/s]
 46%|####5     | 72/157 [00:04<00:04, 17.14it/s]
 47%|####7     | 74/157 [00:04<00:04, 16.98it/s]
 48%|####8     | 76/157 [00:04<00:04, 17.07it/s]
 50%|####9     | 78/157 [00:04<00:04, 17.12it/s]
 51%|#####     | 80/157 [00:04<00:04, 17.03it/s]
 52%|#####2    | 82/157 [00:04<00:04, 17.07it/s]
 54%|#####3    | 84/157 [00:04<00:04, 17.11it/s]
 55%|#####4    | 86/157 [00:05<00:04, 17.17it/s]
 56%|#####6    | 88/157 [00:05<00:04, 17.04it/s]
 57%|#####7    | 90/157 [00:05<00:03, 17.07it/s]
 59%|#####8    | 92/157 [00:05<00:03, 17.11it/s]
 60%|#####9    | 94/157 [00:05<00:03, 17.08it/s]
 61%|######1   | 96/157 [00:05<00:03, 17.04it/s]
 62%|######2   | 98/157 [00:05<00:03, 17.07it/s]
 64%|######3   | 100/157 [00:05<00:03, 17.12it/s]
 65%|######4   | 102/157 [00:06<00:03, 16.98it/s]
 66%|######6   | 104/157 [00:06<00:03, 17.05it/s]
 68%|######7   | 106/157 [00:06<00:02, 17.11it/s]
 69%|######8   | 108/157 [00:06<00:02, 17.09it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.94it/s]
 71%|#######1  | 112/157 [00:06<00:02, 16.96it/s]
 73%|#######2  | 114/157 [00:06<00:02, 16.98it/s]
 74%|#######3  | 116/157 [00:06<00:02, 16.70it/s]
 75%|#######5  | 118/157 [00:06<00:02, 16.69it/s]
 76%|#######6  | 120/157 [00:07<00:02, 16.70it/s]
 78%|#######7  | 122/157 [00:07<00:02, 16.65it/s]
 79%|#######8  | 124/157 [00:07<00:01, 16.61it/s]
 80%|########  | 126/157 [00:07<00:01, 16.66it/s]
 82%|########1 | 128/157 [00:07<00:01, 16.49it/s]
 83%|########2 | 130/157 [00:07<00:01, 16.56it/s]
 84%|########4 | 132/157 [00:07<00:01, 16.55it/s]
 85%|########5 | 134/157 [00:07<00:01, 16.50it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.58it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.64it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.64it/s]
 90%|######### | 142/157 [00:08<00:00, 16.65it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.83it/s]
 93%|#########2| 146/157 [00:08<00:00, 16.99it/s]
 94%|#########4| 148/157 [00:08<00:00, 16.97it/s]
 96%|#########5| 150/157 [00:08<00:00, 16.97it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.98it/s]
 98%|#########8| 154/157 [00:09<00:00, 17.05it/s]
 99%|#########9| 156/157 [00:09<00:00, 17.11it/s]
100%|##########| 157/157 [00:09<00:00, 16.91it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:40, 19.03it/s]
  1%|          | 4/781 [00:00<00:40, 19.19it/s]
  1%|          | 6/781 [00:00<00:40, 19.22it/s]
  1%|1         | 8/781 [00:00<00:40, 19.10it/s]
  1%|1         | 10/781 [00:00<00:40, 19.00it/s]
  2%|1         | 12/781 [00:00<00:40, 19.05it/s]
  2%|1         | 14/781 [00:00<00:40, 18.99it/s]
  2%|2         | 16/781 [00:00<00:40, 18.94it/s]
  2%|2         | 18/781 [00:00<00:40, 18.90it/s]
  3%|2         | 20/781 [00:01<00:40, 19.02it/s]
  3%|2         | 22/781 [00:01<00:39, 19.10it/s]
  3%|3         | 24/781 [00:01<00:39, 19.13it/s]
  3%|3         | 26/781 [00:01<00:39, 19.16it/s]
  4%|3         | 28/781 [00:01<00:39, 18.86it/s]
  4%|3         | 30/781 [00:01<00:39, 18.85it/s]
  4%|4         | 32/781 [00:01<00:39, 18.77it/s]
  4%|4         | 34/781 [00:01<00:39, 18.80it/s]
  5%|4         | 36/781 [00:01<00:39, 18.69it/s]
  5%|4         | 38/781 [00:02<00:39, 18.75it/s]
  5%|5         | 40/781 [00:02<00:39, 18.80it/s]
  5%|5         | 42/781 [00:02<00:39, 18.83it/s]
  6%|5         | 44/781 [00:02<00:39, 18.89it/s]
  6%|5         | 46/781 [00:02<00:38, 18.91it/s]
  6%|6         | 48/781 [00:02<00:38, 18.88it/s]
  6%|6         | 50/781 [00:02<00:38, 18.90it/s]
  7%|6         | 52/781 [00:02<00:38, 18.89it/s]
  7%|6         | 54/781 [00:02<00:38, 18.88it/s]
  7%|7         | 56/781 [00:02<00:38, 18.92it/s]
  7%|7         | 58/781 [00:03<00:38, 18.91it/s]
  8%|7         | 60/781 [00:03<00:38, 18.92it/s]
  8%|7         | 62/781 [00:03<00:38, 18.92it/s]
  8%|8         | 64/781 [00:03<00:37, 18.92it/s]
  8%|8         | 66/781 [00:03<00:37, 18.94it/s]
  9%|8         | 68/781 [00:03<00:37, 19.06it/s]
  9%|8         | 70/781 [00:03<00:37, 19.08it/s]
  9%|9         | 72/781 [00:03<00:37, 19.02it/s]
  9%|9         | 74/781 [00:03<00:37, 19.06it/s]
 10%|9         | 76/781 [00:04<00:37, 19.02it/s]
 10%|9         | 78/781 [00:04<00:36, 19.01it/s]
 10%|#         | 80/781 [00:04<00:36, 19.01it/s]
 10%|#         | 82/781 [00:04<00:36, 19.03it/s]
 11%|#         | 84/781 [00:04<00:36, 19.09it/s]
 11%|#1        | 86/781 [00:04<00:36, 19.20it/s]
 11%|#1        | 88/781 [00:04<00:35, 19.31it/s]
 12%|#1        | 90/781 [00:04<00:35, 19.36it/s]
 12%|#1        | 92/781 [00:04<00:35, 19.29it/s]
 12%|#2        | 94/781 [00:04<00:35, 19.16it/s]
 12%|#2        | 96/781 [00:05<00:35, 19.08it/s]
 13%|#2        | 98/781 [00:05<00:35, 19.06it/s]
 13%|#2        | 100/781 [00:05<00:35, 19.02it/s]
 13%|#3        | 102/781 [00:05<00:35, 19.01it/s]
 13%|#3        | 104/781 [00:05<00:35, 18.98it/s]
 14%|#3        | 106/781 [00:05<00:35, 18.94it/s]
 14%|#3        | 108/781 [00:05<00:35, 18.92it/s]
 14%|#4        | 110/781 [00:05<00:35, 18.92it/s]
 14%|#4        | 112/781 [00:05<00:35, 18.89it/s]
 15%|#4        | 114/781 [00:06<00:35, 18.89it/s]
 15%|#4        | 116/781 [00:06<00:35, 18.85it/s]
 15%|#5        | 118/781 [00:06<00:35, 18.82it/s]
 15%|#5        | 120/781 [00:06<00:34, 18.89it/s]
 16%|#5        | 122/781 [00:06<00:34, 19.03it/s]
 16%|#5        | 124/781 [00:06<00:34, 19.12it/s]
 16%|#6        | 126/781 [00:06<00:34, 19.20it/s]
 16%|#6        | 128/781 [00:06<00:34, 19.16it/s]
 17%|#6        | 130/781 [00:06<00:34, 19.13it/s]
 17%|#6        | 132/781 [00:06<00:34, 19.07it/s]
 17%|#7        | 134/781 [00:07<00:33, 19.15it/s]
 17%|#7        | 136/781 [00:07<00:33, 19.25it/s]
 18%|#7        | 138/781 [00:07<00:33, 19.34it/s]
 18%|#7        | 140/781 [00:07<00:33, 19.38it/s]
 18%|#8        | 142/781 [00:07<00:32, 19.43it/s]
 18%|#8        | 144/781 [00:07<00:32, 19.45it/s]
 19%|#8        | 146/781 [00:07<00:32, 19.49it/s]
 19%|#8        | 148/781 [00:07<00:32, 19.48it/s]
 19%|#9        | 150/781 [00:07<00:32, 19.49it/s]
 19%|#9        | 152/781 [00:07<00:32, 19.50it/s]
 20%|#9        | 154/781 [00:08<00:32, 19.50it/s]
 20%|#9        | 156/781 [00:08<00:32, 19.52it/s]
 20%|##        | 158/781 [00:08<00:31, 19.50it/s]
 20%|##        | 160/781 [00:08<00:31, 19.54it/s]
 21%|##        | 162/781 [00:08<00:31, 19.53it/s]
 21%|##        | 164/781 [00:08<00:31, 19.48it/s]
 21%|##1       | 166/781 [00:08<00:31, 19.45it/s]
 22%|##1       | 168/781 [00:08<00:31, 19.45it/s]
 22%|##1       | 170/781 [00:08<00:31, 19.47it/s]
 22%|##2       | 172/781 [00:09<00:31, 19.50it/s]
 22%|##2       | 174/781 [00:09<00:31, 19.40it/s]
 23%|##2       | 176/781 [00:09<00:31, 19.26it/s]
 23%|##2       | 178/781 [00:09<00:31, 19.18it/s]
 23%|##3       | 180/781 [00:09<00:31, 19.10it/s]
 23%|##3       | 182/781 [00:09<00:31, 19.06it/s]
 24%|##3       | 184/781 [00:09<00:31, 19.01it/s]
 24%|##3       | 186/781 [00:09<00:31, 19.00it/s]
 24%|##4       | 188/781 [00:09<00:31, 19.00it/s]
 24%|##4       | 190/781 [00:09<00:31, 18.96it/s]
 25%|##4       | 192/781 [00:10<00:30, 19.01it/s]
 25%|##4       | 194/781 [00:10<00:30, 19.12it/s]
 25%|##5       | 196/781 [00:10<00:30, 19.21it/s]
 25%|##5       | 198/781 [00:10<00:30, 19.26it/s]
 26%|##5       | 200/781 [00:10<00:30, 19.36it/s]
 26%|##5       | 202/781 [00:10<00:29, 19.41it/s]
 26%|##6       | 204/781 [00:10<00:29, 19.42it/s]
 26%|##6       | 206/781 [00:10<00:29, 19.22it/s]
 27%|##6       | 208/781 [00:10<00:29, 19.29it/s]
 27%|##6       | 210/781 [00:10<00:29, 19.11it/s]
 27%|##7       | 212/781 [00:11<00:29, 19.02it/s]
 27%|##7       | 214/781 [00:11<00:29, 18.97it/s]
 28%|##7       | 216/781 [00:11<00:29, 19.07it/s]
 28%|##7       | 218/781 [00:11<00:30, 18.73it/s]
 28%|##8       | 220/781 [00:11<00:30, 18.66it/s]
 28%|##8       | 222/781 [00:11<00:30, 18.57it/s]
 29%|##8       | 224/781 [00:11<00:29, 18.62it/s]
 29%|##8       | 226/781 [00:11<00:29, 18.61it/s]
 29%|##9       | 228/781 [00:11<00:29, 18.58it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.65it/s]
 30%|##9       | 232/781 [00:12<00:29, 18.72it/s]
 30%|##9       | 234/781 [00:12<00:29, 18.81it/s]
 30%|###       | 236/781 [00:12<00:28, 18.81it/s]
 30%|###       | 238/781 [00:12<00:28, 18.85it/s]
 31%|###       | 240/781 [00:12<00:28, 19.00it/s]
 31%|###       | 242/781 [00:12<00:28, 19.12it/s]
 31%|###1      | 244/781 [00:12<00:28, 19.05it/s]
 31%|###1      | 246/781 [00:12<00:28, 19.03it/s]
 32%|###1      | 248/781 [00:13<00:27, 19.18it/s]
 32%|###2      | 250/781 [00:13<00:27, 19.32it/s]
 32%|###2      | 252/781 [00:13<00:27, 19.27it/s]
 33%|###2      | 254/781 [00:13<00:27, 19.35it/s]
 33%|###2      | 256/781 [00:13<00:26, 19.45it/s]
 33%|###3      | 258/781 [00:13<00:26, 19.48it/s]
 33%|###3      | 260/781 [00:13<00:26, 19.49it/s]
 34%|###3      | 262/781 [00:13<00:26, 19.48it/s]
 34%|###3      | 264/781 [00:13<00:26, 19.49it/s]
 34%|###4      | 266/781 [00:13<00:26, 19.49it/s]
 34%|###4      | 268/781 [00:14<00:26, 19.46it/s]
 35%|###4      | 270/781 [00:14<00:26, 19.47it/s]
 35%|###4      | 272/781 [00:14<00:26, 19.46it/s]
 35%|###5      | 274/781 [00:14<00:26, 19.47it/s]
 35%|###5      | 276/781 [00:14<00:25, 19.48it/s]
 36%|###5      | 278/781 [00:14<00:25, 19.48it/s]
 36%|###5      | 280/781 [00:14<00:25, 19.49it/s]
 36%|###6      | 282/781 [00:14<00:25, 19.48it/s]
 36%|###6      | 284/781 [00:14<00:25, 19.49it/s]
 37%|###6      | 286/781 [00:14<00:25, 19.48it/s]
 37%|###6      | 288/781 [00:15<00:25, 19.47it/s]
 37%|###7      | 290/781 [00:15<00:25, 19.47it/s]
 37%|###7      | 292/781 [00:15<00:25, 19.38it/s]
 38%|###7      | 294/781 [00:15<00:25, 19.41it/s]
 38%|###7      | 296/781 [00:15<00:24, 19.46it/s]
 38%|###8      | 298/781 [00:15<00:24, 19.46it/s]
 38%|###8      | 300/781 [00:15<00:24, 19.42it/s]
 39%|###8      | 302/781 [00:15<00:24, 19.44it/s]
 39%|###8      | 304/781 [00:15<00:24, 19.44it/s]
 39%|###9      | 306/781 [00:15<00:24, 19.35it/s]
 39%|###9      | 308/781 [00:16<00:24, 19.21it/s]
 40%|###9      | 310/781 [00:16<00:24, 19.13it/s]
 40%|###9      | 312/781 [00:16<00:24, 19.10it/s]
 40%|####      | 314/781 [00:16<00:24, 19.14it/s]
 40%|####      | 316/781 [00:16<00:24, 19.24it/s]
 41%|####      | 318/781 [00:16<00:23, 19.34it/s]
 41%|####      | 320/781 [00:16<00:23, 19.38it/s]
 41%|####1     | 322/781 [00:16<00:23, 19.39it/s]
 41%|####1     | 324/781 [00:16<00:23, 19.31it/s]
 42%|####1     | 326/781 [00:17<00:23, 19.33it/s]
 42%|####1     | 328/781 [00:17<00:23, 19.35it/s]
 42%|####2     | 330/781 [00:17<00:23, 19.38it/s]
 43%|####2     | 332/781 [00:17<00:23, 19.37it/s]
 43%|####2     | 334/781 [00:17<00:23, 19.35it/s]
 43%|####3     | 336/781 [00:17<00:23, 19.23it/s]
 43%|####3     | 338/781 [00:17<00:23, 19.16it/s]
 44%|####3     | 340/781 [00:17<00:23, 19.05it/s]
 44%|####3     | 342/781 [00:17<00:23, 19.02it/s]
 44%|####4     | 344/781 [00:17<00:23, 19.00it/s]
 44%|####4     | 346/781 [00:18<00:22, 19.01it/s]
 45%|####4     | 348/781 [00:18<00:22, 18.99it/s]
 45%|####4     | 350/781 [00:18<00:22, 18.94it/s]
 45%|####5     | 352/781 [00:18<00:22, 18.95it/s]
 45%|####5     | 354/781 [00:18<00:22, 18.92it/s]
 46%|####5     | 356/781 [00:18<00:22, 18.98it/s]
 46%|####5     | 358/781 [00:18<00:22, 19.12it/s]
 46%|####6     | 360/781 [00:18<00:21, 19.20it/s]
 46%|####6     | 362/781 [00:18<00:21, 19.27it/s]
 47%|####6     | 364/781 [00:19<00:21, 19.30it/s]
 47%|####6     | 366/781 [00:19<00:21, 19.32it/s]
 47%|####7     | 368/781 [00:19<00:21, 19.30it/s]
 47%|####7     | 370/781 [00:19<00:21, 19.31it/s]
 48%|####7     | 372/781 [00:19<00:21, 19.35it/s]
 48%|####7     | 374/781 [00:19<00:21, 19.35it/s]
 48%|####8     | 376/781 [00:19<00:20, 19.38it/s]
 48%|####8     | 378/781 [00:19<00:20, 19.39it/s]
 49%|####8     | 380/781 [00:19<00:20, 19.40it/s]
 49%|####8     | 382/781 [00:19<00:20, 19.41it/s]
 49%|####9     | 384/781 [00:20<00:20, 19.37it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.35it/s]
 50%|####9     | 388/781 [00:20<00:20, 19.37it/s]
 50%|####9     | 390/781 [00:20<00:20, 19.36it/s]
 50%|#####     | 392/781 [00:20<00:20, 19.37it/s]
 50%|#####     | 394/781 [00:20<00:19, 19.39it/s]
 51%|#####     | 396/781 [00:20<00:19, 19.34it/s]
 51%|#####     | 398/781 [00:20<00:19, 19.29it/s]
 51%|#####1    | 400/781 [00:20<00:19, 19.09it/s]
 51%|#####1    | 402/781 [00:20<00:19, 19.04it/s]
 52%|#####1    | 404/781 [00:21<00:19, 18.85it/s]
 52%|#####1    | 406/781 [00:21<00:19, 18.91it/s]
 52%|#####2    | 408/781 [00:21<00:19, 18.82it/s]
 52%|#####2    | 410/781 [00:21<00:19, 18.82it/s]
 53%|#####2    | 412/781 [00:21<00:19, 18.71it/s]
 53%|#####3    | 414/781 [00:21<00:19, 18.77it/s]
 53%|#####3    | 416/781 [00:21<00:19, 18.78it/s]
 54%|#####3    | 418/781 [00:21<00:19, 18.76it/s]
 54%|#####3    | 420/781 [00:21<00:19, 18.76it/s]
 54%|#####4    | 422/781 [00:22<00:19, 18.67it/s]
 54%|#####4    | 424/781 [00:22<00:19, 18.67it/s]
 55%|#####4    | 426/781 [00:22<00:19, 18.64it/s]
 55%|#####4    | 428/781 [00:22<00:18, 18.77it/s]
 55%|#####5    | 430/781 [00:22<00:18, 18.89it/s]
 55%|#####5    | 432/781 [00:22<00:18, 18.98it/s]
 56%|#####5    | 434/781 [00:22<00:18, 19.07it/s]
 56%|#####5    | 436/781 [00:22<00:17, 19.21it/s]
 56%|#####6    | 438/781 [00:22<00:18, 19.01it/s]
 56%|#####6    | 440/781 [00:22<00:17, 19.04it/s]
 57%|#####6    | 442/781 [00:23<00:17, 19.05it/s]
 57%|#####6    | 444/781 [00:23<00:17, 19.07it/s]
 57%|#####7    | 446/781 [00:23<00:17, 19.01it/s]
 57%|#####7    | 448/781 [00:23<00:17, 19.05it/s]
 58%|#####7    | 450/781 [00:23<00:17, 19.24it/s]
 58%|#####7    | 452/781 [00:23<00:17, 19.02it/s]
 58%|#####8    | 454/781 [00:23<00:17, 19.04it/s]
 58%|#####8    | 456/781 [00:23<00:16, 19.23it/s]
 59%|#####8    | 458/781 [00:23<00:17, 19.00it/s]
 59%|#####8    | 460/781 [00:24<00:16, 18.94it/s]
 59%|#####9    | 462/781 [00:24<00:16, 18.97it/s]
 59%|#####9    | 464/781 [00:24<00:16, 19.04it/s]
 60%|#####9    | 466/781 [00:24<00:16, 19.19it/s]
 60%|#####9    | 468/781 [00:24<00:16, 18.81it/s]
 60%|######    | 470/781 [00:24<00:16, 18.70it/s]
 60%|######    | 472/781 [00:24<00:16, 18.78it/s]
 61%|######    | 474/781 [00:24<00:16, 18.55it/s]
 61%|######    | 476/781 [00:24<00:16, 18.60it/s]
 61%|######1   | 478/781 [00:25<00:16, 18.57it/s]
 61%|######1   | 480/781 [00:25<00:16, 18.61it/s]
 62%|######1   | 482/781 [00:25<00:16, 18.65it/s]
 62%|######1   | 484/781 [00:25<00:15, 18.63it/s]
 62%|######2   | 486/781 [00:25<00:15, 18.62it/s]
 62%|######2   | 488/781 [00:25<00:15, 18.66it/s]
 63%|######2   | 490/781 [00:25<00:15, 18.71it/s]
 63%|######2   | 492/781 [00:25<00:15, 18.89it/s]
 63%|######3   | 494/781 [00:25<00:15, 19.01it/s]
 64%|######3   | 496/781 [00:25<00:15, 18.94it/s]
 64%|######3   | 498/781 [00:26<00:14, 18.99it/s]
 64%|######4   | 500/781 [00:26<00:14, 19.16it/s]
 64%|######4   | 502/781 [00:26<00:14, 18.89it/s]
 65%|######4   | 504/781 [00:26<00:14, 18.85it/s]
 65%|######4   | 506/781 [00:26<00:14, 18.79it/s]
 65%|######5   | 508/781 [00:26<00:14, 18.77it/s]
 65%|######5   | 510/781 [00:26<00:14, 18.93it/s]
 66%|######5   | 512/781 [00:26<00:14, 19.04it/s]
 66%|######5   | 514/781 [00:26<00:14, 18.98it/s]
 66%|######6   | 516/781 [00:27<00:13, 18.93it/s]
 66%|######6   | 518/781 [00:27<00:13, 18.88it/s]
 67%|######6   | 520/781 [00:27<00:13, 18.98it/s]
 67%|######6   | 522/781 [00:27<00:13, 18.74it/s]
 67%|######7   | 524/781 [00:27<00:13, 18.74it/s]
 67%|######7   | 526/781 [00:27<00:13, 18.80it/s]
 68%|######7   | 528/781 [00:27<00:13, 18.94it/s]
 68%|######7   | 530/781 [00:27<00:13, 19.04it/s]
 68%|######8   | 532/781 [00:27<00:13, 19.11it/s]
 68%|######8   | 534/781 [00:27<00:12, 19.15it/s]
 69%|######8   | 536/781 [00:28<00:12, 19.20it/s]
 69%|######8   | 538/781 [00:28<00:12, 19.23it/s]
 69%|######9   | 540/781 [00:28<00:12, 19.04it/s]
 69%|######9   | 542/781 [00:28<00:12, 18.93it/s]
 70%|######9   | 544/781 [00:28<00:12, 18.93it/s]
 70%|######9   | 546/781 [00:28<00:12, 19.02it/s]
 70%|#######   | 548/781 [00:28<00:12, 19.02it/s]
 70%|#######   | 550/781 [00:28<00:12, 19.06it/s]
 71%|#######   | 552/781 [00:28<00:11, 19.30it/s]
 71%|#######   | 554/781 [00:29<00:11, 19.15it/s]
 71%|#######1  | 556/781 [00:29<00:11, 19.14it/s]
 71%|#######1  | 558/781 [00:29<00:11, 19.29it/s]
 72%|#######1  | 560/781 [00:29<00:11, 19.13it/s]
 72%|#######1  | 562/781 [00:29<00:11, 19.13it/s]
 72%|#######2  | 564/781 [00:29<00:11, 19.13it/s]
 72%|#######2  | 566/781 [00:29<00:11, 19.28it/s]
 73%|#######2  | 568/781 [00:29<00:11, 19.03it/s]
 73%|#######2  | 570/781 [00:29<00:11, 18.99it/s]
 73%|#######3  | 572/781 [00:29<00:10, 19.02it/s]
 73%|#######3  | 574/781 [00:30<00:10, 19.11it/s]
 74%|#######3  | 576/781 [00:30<00:10, 19.33it/s]
 74%|#######4  | 578/781 [00:30<00:10, 19.29it/s]
 74%|#######4  | 580/781 [00:30<00:10, 19.10it/s]
 75%|#######4  | 582/781 [00:30<00:10, 19.15it/s]
 75%|#######4  | 584/781 [00:30<00:10, 19.17it/s]
 75%|#######5  | 586/781 [00:30<00:10, 19.18it/s]
 75%|#######5  | 588/781 [00:30<00:10, 18.97it/s]
 76%|#######5  | 590/781 [00:30<00:10, 18.88it/s]
 76%|#######5  | 592/781 [00:31<00:10, 18.87it/s]
 76%|#######6  | 594/781 [00:31<00:09, 19.11it/s]
 76%|#######6  | 596/781 [00:31<00:09, 18.87it/s]
 77%|#######6  | 598/781 [00:31<00:09, 19.08it/s]
 77%|#######6  | 600/781 [00:31<00:09, 18.83it/s]
 77%|#######7  | 602/781 [00:31<00:09, 18.63it/s]
 77%|#######7  | 604/781 [00:31<00:09, 18.64it/s]
 78%|#######7  | 606/781 [00:31<00:09, 18.81it/s]
 78%|#######7  | 608/781 [00:31<00:09, 18.89it/s]
 78%|#######8  | 610/781 [00:31<00:09, 18.94it/s]
 78%|#######8  | 612/781 [00:32<00:08, 19.02it/s]
 79%|#######8  | 614/781 [00:32<00:08, 19.05it/s]
 79%|#######8  | 616/781 [00:32<00:08, 19.12it/s]
 79%|#######9  | 618/781 [00:32<00:08, 19.14it/s]
 79%|#######9  | 620/781 [00:32<00:08, 19.19it/s]
 80%|#######9  | 622/781 [00:32<00:08, 19.33it/s]
 80%|#######9  | 624/781 [00:32<00:08, 19.12it/s]
 80%|########  | 626/781 [00:32<00:08, 18.99it/s]
 80%|########  | 628/781 [00:32<00:08, 19.05it/s]
 81%|########  | 630/781 [00:33<00:07, 19.27it/s]
 81%|########  | 632/781 [00:33<00:07, 19.10it/s]
 81%|########1 | 634/781 [00:33<00:07, 19.12it/s]
 81%|########1 | 636/781 [00:33<00:07, 19.12it/s]
 82%|########1 | 638/781 [00:33<00:07, 19.03it/s]
 82%|########1 | 640/781 [00:33<00:07, 18.94it/s]
 82%|########2 | 642/781 [00:33<00:07, 18.81it/s]
 82%|########2 | 644/781 [00:33<00:07, 18.91it/s]
 83%|########2 | 646/781 [00:33<00:07, 18.81it/s]
 83%|########2 | 648/781 [00:33<00:07, 18.57it/s]
 83%|########3 | 650/781 [00:34<00:07, 18.56it/s]
 83%|########3 | 652/781 [00:34<00:06, 18.56it/s]
 84%|########3 | 654/781 [00:34<00:06, 18.61it/s]
 84%|########3 | 656/781 [00:34<00:06, 18.67it/s]
 84%|########4 | 658/781 [00:34<00:06, 18.99it/s]
 85%|########4 | 660/781 [00:34<00:06, 18.91it/s]
 85%|########4 | 662/781 [00:34<00:06, 19.17it/s]
 85%|########5 | 664/781 [00:34<00:06, 19.18it/s]
 85%|########5 | 666/781 [00:34<00:06, 19.02it/s]
 86%|########5 | 668/781 [00:35<00:05, 19.07it/s]
 86%|########5 | 670/781 [00:35<00:05, 19.14it/s]
 86%|########6 | 672/781 [00:35<00:05, 19.09it/s]
 86%|########6 | 674/781 [00:35<00:05, 19.15it/s]
 87%|########6 | 676/781 [00:35<00:05, 19.21it/s]
 87%|########6 | 678/781 [00:35<00:05, 19.28it/s]
 87%|########7 | 680/781 [00:35<00:05, 19.31it/s]
 87%|########7 | 682/781 [00:35<00:05, 19.33it/s]
 88%|########7 | 684/781 [00:35<00:05, 19.33it/s]
 88%|########7 | 686/781 [00:35<00:04, 19.34it/s]
 88%|########8 | 688/781 [00:36<00:04, 19.37it/s]
 88%|########8 | 690/781 [00:36<00:04, 19.31it/s]
 89%|########8 | 692/781 [00:36<00:04, 19.35it/s]
 89%|########8 | 694/781 [00:36<00:04, 19.33it/s]
 89%|########9 | 696/781 [00:36<00:04, 19.29it/s]
 89%|########9 | 698/781 [00:36<00:04, 19.28it/s]
 90%|########9 | 700/781 [00:36<00:04, 19.29it/s]
 90%|########9 | 702/781 [00:36<00:04, 19.23it/s]
 90%|######### | 704/781 [00:36<00:04, 19.23it/s]
 90%|######### | 706/781 [00:36<00:03, 19.24it/s]
 91%|######### | 708/781 [00:37<00:03, 19.25it/s]
 91%|######### | 710/781 [00:37<00:03, 19.30it/s]
 91%|#########1| 712/781 [00:37<00:03, 19.35it/s]
 91%|#########1| 714/781 [00:37<00:03, 19.37it/s]
 92%|#########1| 716/781 [00:37<00:03, 19.41it/s]
 92%|#########1| 718/781 [00:37<00:03, 19.40it/s]
 92%|#########2| 720/781 [00:37<00:03, 19.36it/s]
 92%|#########2| 722/781 [00:37<00:03, 19.31it/s]
 93%|#########2| 724/781 [00:37<00:02, 19.30it/s]
 93%|#########2| 726/781 [00:38<00:02, 19.31it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.33it/s]
 93%|#########3| 730/781 [00:38<00:02, 19.34it/s]
 94%|#########3| 732/781 [00:38<00:02, 19.34it/s]
 94%|#########3| 734/781 [00:38<00:02, 19.34it/s]
 94%|#########4| 736/781 [00:38<00:02, 19.37it/s]
 94%|#########4| 738/781 [00:38<00:02, 19.38it/s]
 95%|#########4| 740/781 [00:38<00:02, 19.43it/s]
 95%|#########5| 742/781 [00:38<00:02, 19.43it/s]
 95%|#########5| 744/781 [00:38<00:01, 19.48it/s]
 96%|#########5| 746/781 [00:39<00:01, 19.46it/s]
 96%|#########5| 748/781 [00:39<00:01, 19.45it/s]
 96%|#########6| 750/781 [00:39<00:01, 19.44it/s]
 96%|#########6| 752/781 [00:39<00:01, 19.45it/s]
 97%|#########6| 754/781 [00:39<00:01, 19.45it/s]
 97%|#########6| 756/781 [00:39<00:01, 19.46it/s]
 97%|#########7| 758/781 [00:39<00:01, 19.47it/s]
 97%|#########7| 760/781 [00:39<00:01, 19.48it/s]
 98%|#########7| 762/781 [00:39<00:00, 19.48it/s]
 98%|#########7| 764/781 [00:39<00:00, 19.47it/s]
 98%|#########8| 766/781 [00:40<00:00, 19.49it/s]
 98%|#########8| 768/781 [00:40<00:00, 19.49it/s]
 99%|#########8| 770/781 [00:40<00:00, 19.48it/s]
 99%|#########8| 772/781 [00:40<00:00, 19.38it/s]
 99%|#########9| 774/781 [00:40<00:00, 19.30it/s]
 99%|#########9| 776/781 [00:40<00:00, 19.25it/s]
100%|#########9| 778/781 [00:40<00:00, 19.33it/s]
100%|#########9| 780/781 [00:40<00:00, 19.23it/s]
100%|##########| 781/781 [00:40<00:00, 19.12it/s]
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:02<01:31,  2.14s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:32,  2.14s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:32,  2.16s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:32,  2.15s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:32,  2.16s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:32,  2.16s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:32,  2.16s/it]
Validation Progress:   2%|2         | 1/44 [00:02<01:34,  2.20s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.53s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.53s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.54s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.54s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.54s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.54s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.54s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:07,  1.60s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.45s/it]
Validation Progress:   7%|6         | 3/44 [00:04<01:00,  1.48s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.33s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.33s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.34s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.34s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.34s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.34s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:53,  1.34s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:54,  1.36s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:48,  1.25s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:48,  1.25s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:48,  1.26s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:48,  1.25s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:48,  1.25s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:48,  1.26s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:49,  1.26s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:49,  1.27s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:45,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:46,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:46,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:46,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:46,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:46,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:46,  1.21s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:46,  1.22s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.20s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:44,  1.21s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:43,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:43,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:43,  1.20s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:43,  1.20s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:42,  1.21s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.22s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.28s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.30s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:56<00:00,  1.28s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.00s/it]
100%|##########| 1/1 [00:03<00:00,  3.00s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.81it/s]
  3%|2         | 4/157 [00:00<00:10, 14.53it/s]
  4%|3         | 6/157 [00:00<00:10, 15.03it/s]
  5%|5         | 8/157 [00:00<00:09, 15.27it/s]
  6%|6         | 10/157 [00:00<00:09, 15.53it/s]
  8%|7         | 12/157 [00:00<00:09, 15.75it/s]
  9%|8         | 14/157 [00:00<00:08, 15.89it/s]
 10%|#         | 16/157 [00:01<00:08, 16.03it/s]
 11%|#1        | 18/157 [00:01<00:08, 16.11it/s]
 13%|#2        | 20/157 [00:01<00:08, 16.15it/s]
 14%|#4        | 22/157 [00:01<00:08, 16.20it/s]
 15%|#5        | 24/157 [00:01<00:08, 15.95it/s]
 17%|#6        | 26/157 [00:01<00:09, 13.26it/s]
 18%|#7        | 28/157 [00:01<00:09, 14.06it/s]
 19%|#9        | 30/157 [00:01<00:08, 14.66it/s]
 20%|##        | 32/157 [00:02<00:08, 15.12it/s]
 22%|##1       | 34/157 [00:02<00:07, 15.45it/s]
 23%|##2       | 36/157 [00:02<00:07, 15.70it/s]
 24%|##4       | 38/157 [00:02<00:07, 15.86it/s]
 25%|##5       | 40/157 [00:02<00:07, 15.97it/s]
 27%|##6       | 42/157 [00:02<00:07, 16.09it/s]
 28%|##8       | 44/157 [00:02<00:07, 16.11it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.05it/s]
 31%|###       | 48/157 [00:03<00:06, 16.01it/s]
 32%|###1      | 50/157 [00:03<00:06, 15.98it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.00it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.08it/s]
 36%|###5      | 56/157 [00:03<00:06, 16.15it/s]
 37%|###6      | 58/157 [00:03<00:06, 16.09it/s]
 38%|###8      | 60/157 [00:03<00:06, 16.06it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.13it/s]
 41%|####      | 64/157 [00:04<00:05, 16.20it/s]
 42%|####2     | 66/157 [00:04<00:05, 16.24it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.28it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.29it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.31it/s]
 47%|####7     | 74/157 [00:04<00:05, 16.33it/s]
 48%|####8     | 76/157 [00:04<00:04, 16.34it/s]
 50%|####9     | 78/157 [00:04<00:04, 16.34it/s]
 51%|#####     | 80/157 [00:05<00:04, 16.36it/s]
 52%|#####2    | 82/157 [00:05<00:04, 16.35it/s]
 54%|#####3    | 84/157 [00:05<00:04, 16.35it/s]
 55%|#####4    | 86/157 [00:05<00:04, 16.35it/s]
 56%|#####6    | 88/157 [00:05<00:04, 16.33it/s]
 57%|#####7    | 90/157 [00:05<00:04, 16.34it/s]
 59%|#####8    | 92/157 [00:05<00:03, 16.34it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.36it/s]
 61%|######1   | 96/157 [00:06<00:03, 16.35it/s]
 62%|######2   | 98/157 [00:06<00:03, 16.36it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.35it/s]
 65%|######4   | 102/157 [00:06<00:03, 16.34it/s]
 66%|######6   | 104/157 [00:06<00:03, 16.34it/s]
 68%|######7   | 106/157 [00:06<00:03, 16.35it/s]
 69%|######8   | 108/157 [00:06<00:02, 16.35it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.32it/s]
 71%|#######1  | 112/157 [00:07<00:02, 16.27it/s]
 73%|#######2  | 114/157 [00:07<00:02, 16.25it/s]
 74%|#######3  | 116/157 [00:07<00:02, 16.22it/s]
 75%|#######5  | 118/157 [00:07<00:02, 16.20it/s]
 76%|#######6  | 120/157 [00:07<00:02, 16.20it/s]
 78%|#######7  | 122/157 [00:07<00:02, 16.23it/s]
 79%|#######8  | 124/157 [00:07<00:02, 16.22it/s]
 80%|########  | 126/157 [00:07<00:01, 16.20it/s]
 82%|########1 | 128/157 [00:08<00:01, 16.24it/s]
 83%|########2 | 130/157 [00:08<00:01, 16.24it/s]
 84%|########4 | 132/157 [00:08<00:01, 16.23it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.24it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.25it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.25it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.26it/s]
 90%|######### | 142/157 [00:08<00:00, 16.26it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.28it/s]
 93%|#########2| 146/157 [00:09<00:00, 16.28it/s]
 94%|#########4| 148/157 [00:09<00:00, 16.24it/s]
 96%|#########5| 150/157 [00:09<00:00, 16.22it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.15it/s]
 98%|#########8| 154/157 [00:09<00:00, 16.15it/s]
 99%|#########9| 156/157 [00:09<00:00, 16.19it/s]
100%|##########| 157/157 [00:09<00:00, 16.08it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:43, 17.92it/s]
  1%|          | 4/781 [00:00<00:44, 17.61it/s]
  1%|          | 6/781 [00:00<00:44, 17.54it/s]
  1%|1         | 8/781 [00:00<00:43, 17.68it/s]
  1%|1         | 10/781 [00:00<00:43, 17.81it/s]
  2%|1         | 12/781 [00:00<00:43, 17.86it/s]
  2%|1         | 14/781 [00:00<00:42, 17.92it/s]
  2%|2         | 16/781 [00:00<00:42, 17.92it/s]
  2%|2         | 18/781 [00:01<00:42, 17.94it/s]
  3%|2         | 20/781 [00:01<00:42, 17.89it/s]
  3%|2         | 22/781 [00:01<00:42, 17.92it/s]
  3%|3         | 24/781 [00:01<00:42, 17.98it/s]
  3%|3         | 26/781 [00:01<00:41, 17.98it/s]
  4%|3         | 28/781 [00:01<00:41, 17.99it/s]
  4%|3         | 30/781 [00:01<00:41, 17.99it/s]
  4%|4         | 32/781 [00:01<00:41, 17.98it/s]
  4%|4         | 34/781 [00:01<00:41, 18.00it/s]
  5%|4         | 36/781 [00:02<00:41, 17.99it/s]
  5%|4         | 38/781 [00:02<00:41, 18.02it/s]
  5%|5         | 40/781 [00:02<00:41, 18.05it/s]
  5%|5         | 42/781 [00:02<00:40, 18.04it/s]
  6%|5         | 44/781 [00:02<00:40, 18.04it/s]
  6%|5         | 46/781 [00:02<00:40, 18.03it/s]
  6%|6         | 48/781 [00:02<00:40, 18.02it/s]
  6%|6         | 50/781 [00:02<00:40, 18.01it/s]
  7%|6         | 52/781 [00:02<00:40, 17.96it/s]
  7%|6         | 54/781 [00:03<00:40, 17.86it/s]
  7%|7         | 56/781 [00:03<00:40, 17.77it/s]
  7%|7         | 58/781 [00:03<00:40, 17.86it/s]
  8%|7         | 60/781 [00:03<00:40, 17.89it/s]
  8%|7         | 62/781 [00:03<00:40, 17.95it/s]
  8%|8         | 64/781 [00:03<00:39, 17.98it/s]
  8%|8         | 66/781 [00:03<00:39, 17.99it/s]
  9%|8         | 68/781 [00:03<00:39, 18.00it/s]
  9%|8         | 70/781 [00:03<00:40, 17.71it/s]
  9%|9         | 72/781 [00:04<00:39, 17.76it/s]
  9%|9         | 74/781 [00:04<00:39, 17.75it/s]
 10%|9         | 76/781 [00:04<00:39, 17.70it/s]
 10%|9         | 78/781 [00:04<00:39, 17.59it/s]
 10%|#         | 80/781 [00:04<00:39, 17.54it/s]
 10%|#         | 82/781 [00:04<00:39, 17.48it/s]
 11%|#         | 84/781 [00:04<00:39, 17.51it/s]
 11%|#1        | 86/781 [00:04<00:39, 17.58it/s]
 11%|#1        | 88/781 [00:04<00:39, 17.68it/s]
 12%|#1        | 90/781 [00:05<00:38, 17.76it/s]
 12%|#1        | 92/781 [00:05<00:38, 17.83it/s]
 12%|#2        | 94/781 [00:05<00:38, 17.87it/s]
 12%|#2        | 96/781 [00:05<00:38, 17.88it/s]
 13%|#2        | 98/781 [00:05<00:38, 17.88it/s]
 13%|#2        | 100/781 [00:05<00:37, 17.94it/s]
 13%|#3        | 102/781 [00:05<00:37, 17.92it/s]
 13%|#3        | 104/781 [00:05<00:37, 17.88it/s]
 14%|#3        | 106/781 [00:05<00:37, 17.93it/s]
 14%|#3        | 108/781 [00:06<00:37, 17.96it/s]
 14%|#4        | 110/781 [00:06<00:37, 17.97it/s]
 14%|#4        | 112/781 [00:06<00:37, 17.97it/s]
 15%|#4        | 114/781 [00:06<00:37, 17.97it/s]
 15%|#4        | 116/781 [00:06<00:37, 17.97it/s]
 15%|#5        | 118/781 [00:06<00:36, 17.97it/s]
 15%|#5        | 120/781 [00:06<00:36, 18.01it/s]
 16%|#5        | 122/781 [00:06<00:36, 17.98it/s]
 16%|#5        | 124/781 [00:06<00:36, 17.97it/s]
 16%|#6        | 126/781 [00:07<00:36, 17.97it/s]
 16%|#6        | 128/781 [00:07<00:36, 17.97it/s]
 17%|#6        | 130/781 [00:07<00:36, 17.99it/s]
 17%|#6        | 132/781 [00:07<00:36, 17.96it/s]
 17%|#7        | 134/781 [00:07<00:35, 17.99it/s]
 17%|#7        | 136/781 [00:07<00:35, 18.00it/s]
 18%|#7        | 138/781 [00:07<00:35, 18.04it/s]
 18%|#7        | 140/781 [00:07<00:35, 18.02it/s]
 18%|#8        | 142/781 [00:07<00:35, 18.04it/s]
 18%|#8        | 144/781 [00:08<00:35, 18.02it/s]
 19%|#8        | 146/781 [00:08<00:35, 18.01it/s]
 19%|#8        | 148/781 [00:08<00:35, 17.99it/s]
 19%|#9        | 150/781 [00:08<00:35, 17.97it/s]
 19%|#9        | 152/781 [00:08<00:35, 17.95it/s]
 20%|#9        | 154/781 [00:08<00:34, 17.95it/s]
 20%|#9        | 156/781 [00:08<00:34, 17.96it/s]
 20%|##        | 158/781 [00:08<00:34, 17.99it/s]
 20%|##        | 160/781 [00:08<00:34, 17.97it/s]
 21%|##        | 162/781 [00:09<00:34, 17.96it/s]
 21%|##        | 164/781 [00:09<00:34, 17.98it/s]
 21%|##1       | 166/781 [00:09<00:34, 17.95it/s]
 22%|##1       | 168/781 [00:09<00:34, 17.90it/s]
 22%|##1       | 170/781 [00:09<00:34, 17.88it/s]
 22%|##2       | 172/781 [00:09<00:34, 17.85it/s]
 22%|##2       | 174/781 [00:09<00:33, 17.87it/s]
 23%|##2       | 176/781 [00:09<00:34, 17.79it/s]
 23%|##2       | 178/781 [00:09<00:33, 17.80it/s]
 23%|##3       | 180/781 [00:10<00:33, 17.77it/s]
 23%|##3       | 182/781 [00:10<00:33, 17.83it/s]
 24%|##3       | 184/781 [00:10<00:33, 17.65it/s]
 24%|##3       | 186/781 [00:10<00:33, 17.62it/s]
 24%|##4       | 188/781 [00:10<00:33, 17.57it/s]
 24%|##4       | 190/781 [00:10<00:33, 17.70it/s]
 25%|##4       | 192/781 [00:10<00:33, 17.74it/s]
 25%|##4       | 194/781 [00:10<00:32, 17.80it/s]
 25%|##5       | 196/781 [00:10<00:32, 17.77it/s]
 25%|##5       | 198/781 [00:11<00:32, 17.83it/s]
 26%|##5       | 200/781 [00:11<00:32, 17.88it/s]
 26%|##5       | 202/781 [00:11<00:32, 17.88it/s]
 26%|##6       | 204/781 [00:11<00:32, 17.92it/s]
 26%|##6       | 206/781 [00:11<00:32, 17.95it/s]
 27%|##6       | 208/781 [00:11<00:31, 17.96it/s]
 27%|##6       | 210/781 [00:11<00:31, 17.96it/s]
 27%|##7       | 212/781 [00:11<00:31, 17.96it/s]
 27%|##7       | 214/781 [00:11<00:31, 17.96it/s]
 28%|##7       | 216/781 [00:12<00:31, 17.96it/s]
 28%|##7       | 218/781 [00:12<00:31, 17.96it/s]
 28%|##8       | 220/781 [00:12<00:31, 17.96it/s]
 28%|##8       | 222/781 [00:12<00:31, 17.99it/s]
 29%|##8       | 224/781 [00:12<00:31, 17.95it/s]
 29%|##8       | 226/781 [00:12<00:30, 17.98it/s]
 29%|##9       | 228/781 [00:12<00:30, 17.97it/s]
 29%|##9       | 230/781 [00:12<00:30, 17.97it/s]
 30%|##9       | 232/781 [00:12<00:30, 17.98it/s]
 30%|##9       | 234/781 [00:13<00:30, 18.00it/s]
 30%|###       | 236/781 [00:13<00:30, 18.01it/s]
 30%|###       | 238/781 [00:13<00:30, 18.01it/s]
 31%|###       | 240/781 [00:13<00:30, 18.01it/s]
 31%|###       | 242/781 [00:13<00:30, 17.96it/s]
 31%|###1      | 244/781 [00:13<00:29, 17.92it/s]
 31%|###1      | 246/781 [00:13<00:30, 17.83it/s]
 32%|###1      | 248/781 [00:13<00:30, 17.74it/s]
 32%|###2      | 250/781 [00:13<00:29, 17.77it/s]
 32%|###2      | 252/781 [00:14<00:29, 17.86it/s]
 33%|###2      | 254/781 [00:14<00:29, 17.93it/s]
 33%|###2      | 256/781 [00:14<00:29, 17.95it/s]
 33%|###3      | 258/781 [00:14<00:29, 17.99it/s]
 33%|###3      | 260/781 [00:14<00:28, 17.97it/s]
 34%|###3      | 262/781 [00:14<00:28, 17.97it/s]
 34%|###3      | 264/781 [00:14<00:28, 17.96it/s]
 34%|###4      | 266/781 [00:14<00:28, 17.98it/s]
 34%|###4      | 268/781 [00:14<00:28, 17.99it/s]
 35%|###4      | 270/781 [00:15<00:28, 17.98it/s]
 35%|###4      | 272/781 [00:15<00:28, 17.98it/s]
 35%|###5      | 274/781 [00:15<00:28, 17.97it/s]
 35%|###5      | 276/781 [00:15<00:28, 17.85it/s]
 36%|###5      | 278/781 [00:15<00:28, 17.88it/s]
 36%|###5      | 280/781 [00:15<00:28, 17.81it/s]
 36%|###6      | 282/781 [00:15<00:28, 17.77it/s]
 36%|###6      | 284/781 [00:15<00:27, 17.88it/s]
 37%|###6      | 286/781 [00:15<00:27, 17.93it/s]
 37%|###6      | 288/781 [00:16<00:27, 17.96it/s]
 37%|###7      | 290/781 [00:16<00:27, 17.98it/s]
 37%|###7      | 292/781 [00:16<00:27, 17.98it/s]
 38%|###7      | 294/781 [00:16<00:27, 18.02it/s]
 38%|###7      | 296/781 [00:16<00:26, 18.02it/s]
 38%|###8      | 298/781 [00:16<00:26, 18.01it/s]
 38%|###8      | 300/781 [00:16<00:26, 18.00it/s]
 39%|###8      | 302/781 [00:16<00:26, 17.99it/s]
 39%|###8      | 304/781 [00:16<00:26, 17.84it/s]
 39%|###9      | 306/781 [00:17<00:26, 17.85it/s]
 39%|###9      | 308/781 [00:17<00:26, 17.82it/s]
 40%|###9      | 310/781 [00:17<00:26, 17.81it/s]
 40%|###9      | 312/781 [00:17<00:26, 17.74it/s]
 40%|####      | 314/781 [00:17<00:26, 17.71it/s]
 40%|####      | 316/781 [00:17<00:26, 17.76it/s]
 41%|####      | 318/781 [00:17<00:25, 17.82it/s]
 41%|####      | 320/781 [00:17<00:25, 17.81it/s]
 41%|####1     | 322/781 [00:17<00:25, 17.90it/s]
 41%|####1     | 324/781 [00:18<00:25, 18.00it/s]
 42%|####1     | 326/781 [00:18<00:25, 18.08it/s]
 42%|####1     | 328/781 [00:18<00:24, 18.14it/s]
 42%|####2     | 330/781 [00:18<00:24, 18.16it/s]
 43%|####2     | 332/781 [00:18<00:24, 18.19it/s]
 43%|####2     | 334/781 [00:18<00:24, 18.24it/s]
 43%|####3     | 336/781 [00:18<00:24, 18.27it/s]
 43%|####3     | 338/781 [00:18<00:24, 18.28it/s]
 44%|####3     | 340/781 [00:18<00:24, 18.30it/s]
 44%|####3     | 342/781 [00:19<00:24, 18.28it/s]
 44%|####4     | 344/781 [00:19<00:23, 18.29it/s]
 44%|####4     | 346/781 [00:19<00:23, 18.26it/s]
 45%|####4     | 348/781 [00:19<00:23, 18.17it/s]
 45%|####4     | 350/781 [00:19<00:23, 17.98it/s]
 45%|####5     | 352/781 [00:19<00:23, 18.04it/s]
 45%|####5     | 354/781 [00:19<00:23, 17.99it/s]
 46%|####5     | 356/781 [00:19<00:23, 17.85it/s]
 46%|####5     | 358/781 [00:19<00:23, 17.86it/s]
 46%|####6     | 360/781 [00:20<00:23, 17.95it/s]
 46%|####6     | 362/781 [00:20<00:23, 17.89it/s]
 47%|####6     | 364/781 [00:20<00:23, 17.89it/s]
 47%|####6     | 366/781 [00:20<00:23, 17.89it/s]
 47%|####7     | 368/781 [00:20<00:22, 18.13it/s]
 47%|####7     | 370/781 [00:20<00:22, 18.07it/s]
 48%|####7     | 372/781 [00:20<00:22, 18.05it/s]
 48%|####7     | 374/781 [00:20<00:22, 18.06it/s]
 48%|####8     | 376/781 [00:20<00:22, 18.08it/s]
 48%|####8     | 378/781 [00:21<00:22, 18.12it/s]
 49%|####8     | 380/781 [00:21<00:22, 18.18it/s]
 49%|####8     | 382/781 [00:21<00:21, 18.17it/s]
 49%|####9     | 384/781 [00:21<00:21, 18.17it/s]
 49%|####9     | 386/781 [00:21<00:21, 18.34it/s]
 50%|####9     | 388/781 [00:21<00:21, 18.12it/s]
 50%|####9     | 390/781 [00:21<00:21, 18.19it/s]
 50%|#####     | 392/781 [00:21<00:21, 18.17it/s]
 50%|#####     | 394/781 [00:21<00:20, 18.45it/s]
 51%|#####     | 396/781 [00:22<00:21, 18.26it/s]
 51%|#####     | 398/781 [00:22<00:21, 18.09it/s]
 51%|#####1    | 400/781 [00:22<00:21, 17.92it/s]
 51%|#####1    | 402/781 [00:22<00:21, 17.82it/s]
 52%|#####1    | 404/781 [00:22<00:21, 17.77it/s]
 52%|#####1    | 406/781 [00:22<00:21, 17.76it/s]
 52%|#####2    | 408/781 [00:22<00:20, 18.10it/s]
 52%|#####2    | 410/781 [00:22<00:20, 18.17it/s]
 53%|#####2    | 412/781 [00:22<00:20, 18.06it/s]
 53%|#####3    | 414/781 [00:23<00:20, 18.06it/s]
 53%|#####3    | 416/781 [00:23<00:19, 18.28it/s]
 54%|#####3    | 418/781 [00:23<00:19, 18.22it/s]
 54%|#####3    | 420/781 [00:23<00:19, 18.39it/s]
 54%|#####4    | 422/781 [00:23<00:19, 18.28it/s]
 54%|#####4    | 424/781 [00:23<00:19, 18.21it/s]
 55%|#####4    | 426/781 [00:23<00:19, 18.20it/s]
 55%|#####4    | 428/781 [00:23<00:19, 18.36it/s]
 55%|#####5    | 430/781 [00:23<00:19, 18.34it/s]
 55%|#####5    | 432/781 [00:24<00:18, 18.48it/s]
 56%|#####5    | 434/781 [00:24<00:18, 18.47it/s]
 56%|#####5    | 436/781 [00:24<00:18, 18.43it/s]
 56%|#####6    | 438/781 [00:24<00:18, 18.17it/s]
 56%|#####6    | 440/781 [00:24<00:18, 18.03it/s]
 57%|#####6    | 442/781 [00:24<00:18, 18.35it/s]
 57%|#####6    | 444/781 [00:24<00:18, 18.41it/s]
 57%|#####7    | 446/781 [00:24<00:18, 18.30it/s]
 57%|#####7    | 448/781 [00:24<00:18, 18.26it/s]
 58%|#####7    | 450/781 [00:25<00:17, 18.66it/s]
 58%|#####7    | 452/781 [00:25<00:17, 18.92it/s]
 58%|#####8    | 454/781 [00:25<00:17, 18.48it/s]
 58%|#####8    | 456/781 [00:25<00:17, 18.62it/s]
 59%|#####8    | 458/781 [00:25<00:17, 18.89it/s]
 59%|#####8    | 460/781 [00:25<00:16, 19.01it/s]
 59%|#####9    | 462/781 [00:25<00:16, 19.08it/s]
 59%|#####9    | 464/781 [00:25<00:16, 18.85it/s]
 60%|#####9    | 466/781 [00:25<00:16, 18.70it/s]
 60%|#####9    | 468/781 [00:25<00:17, 18.35it/s]
 60%|######    | 470/781 [00:26<00:17, 18.26it/s]
 60%|######    | 472/781 [00:26<00:16, 18.45it/s]
 61%|######    | 474/781 [00:26<00:16, 18.87it/s]
 61%|######    | 476/781 [00:26<00:15, 19.20it/s]
 61%|######1   | 478/781 [00:26<00:15, 19.42it/s]
 62%|######1   | 481/781 [00:26<00:15, 19.64it/s]
 62%|######1   | 483/781 [00:26<00:15, 19.69it/s]
 62%|######2   | 485/781 [00:26<00:14, 19.77it/s]
 62%|######2   | 488/781 [00:27<00:14, 19.85it/s]
 63%|######2   | 490/781 [00:27<00:14, 19.86it/s]
 63%|######2   | 492/781 [00:27<00:14, 19.89it/s]
 63%|######3   | 494/781 [00:27<00:14, 19.91it/s]
 64%|######3   | 496/781 [00:27<00:14, 19.89it/s]
 64%|######3   | 498/781 [00:27<00:14, 19.89it/s]
 64%|######4   | 500/781 [00:27<00:14, 19.90it/s]
 64%|######4   | 503/781 [00:27<00:13, 19.98it/s]
 65%|######4   | 505/781 [00:27<00:13, 19.97it/s]
 65%|######4   | 507/781 [00:27<00:13, 19.97it/s]
 65%|######5   | 509/781 [00:28<00:13, 19.95it/s]
 65%|######5   | 511/781 [00:28<00:13, 19.91it/s]
 66%|######5   | 513/781 [00:28<00:13, 19.91it/s]
 66%|######5   | 515/781 [00:28<00:13, 19.90it/s]
 66%|######6   | 517/781 [00:28<00:13, 19.90it/s]
 66%|######6   | 519/781 [00:28<00:13, 19.87it/s]
 67%|######6   | 521/781 [00:28<00:13, 19.90it/s]
 67%|######6   | 523/781 [00:28<00:12, 19.90it/s]
 67%|######7   | 525/781 [00:28<00:12, 19.90it/s]
 67%|######7   | 527/781 [00:28<00:12, 19.92it/s]
 68%|######7   | 530/781 [00:29<00:12, 19.97it/s]
 68%|######8   | 532/781 [00:29<00:12, 19.93it/s]
 68%|######8   | 534/781 [00:29<00:12, 19.74it/s]
 69%|######8   | 536/781 [00:29<00:12, 19.62it/s]
 69%|######8   | 538/781 [00:29<00:12, 19.50it/s]
 69%|######9   | 540/781 [00:29<00:12, 19.51it/s]
 69%|######9   | 542/781 [00:29<00:12, 19.49it/s]
 70%|######9   | 544/781 [00:29<00:12, 19.57it/s]
 70%|######9   | 546/781 [00:29<00:12, 19.50it/s]
 70%|#######   | 548/781 [00:30<00:11, 19.60it/s]
 70%|#######   | 550/781 [00:30<00:11, 19.61it/s]
 71%|#######   | 552/781 [00:30<00:11, 19.54it/s]
 71%|#######   | 554/781 [00:30<00:11, 19.44it/s]
 71%|#######1  | 556/781 [00:30<00:11, 19.47it/s]
 71%|#######1  | 558/781 [00:30<00:11, 19.56it/s]
 72%|#######1  | 560/781 [00:30<00:11, 19.62it/s]
 72%|#######1  | 562/781 [00:30<00:11, 19.68it/s]
 72%|#######2  | 564/781 [00:30<00:11, 19.66it/s]
 72%|#######2  | 566/781 [00:30<00:10, 19.65it/s]
 73%|#######2  | 568/781 [00:31<00:10, 19.72it/s]
 73%|#######2  | 570/781 [00:31<00:10, 19.74it/s]
 73%|#######3  | 572/781 [00:31<00:10, 19.80it/s]
 73%|#######3  | 574/781 [00:31<00:10, 19.78it/s]
 74%|#######3  | 576/781 [00:31<00:10, 19.77it/s]
 74%|#######4  | 578/781 [00:31<00:10, 19.72it/s]
 74%|#######4  | 580/781 [00:31<00:10, 19.65it/s]
 75%|#######4  | 582/781 [00:31<00:10, 19.72it/s]
 75%|#######4  | 584/781 [00:31<00:09, 19.75it/s]
 75%|#######5  | 586/781 [00:31<00:09, 19.80it/s]
 75%|#######5  | 588/781 [00:32<00:09, 19.83it/s]
 76%|#######5  | 590/781 [00:32<00:09, 19.81it/s]
 76%|#######5  | 592/781 [00:32<00:09, 19.80it/s]
 76%|#######6  | 594/781 [00:32<00:09, 19.82it/s]
 76%|#######6  | 596/781 [00:32<00:09, 19.82it/s]
 77%|#######6  | 598/781 [00:32<00:09, 19.82it/s]
 77%|#######6  | 600/781 [00:32<00:09, 19.82it/s]
 77%|#######7  | 602/781 [00:32<00:09, 19.72it/s]
 77%|#######7  | 604/781 [00:32<00:08, 19.76it/s]
 78%|#######7  | 606/781 [00:32<00:08, 19.81it/s]
 78%|#######7  | 608/781 [00:33<00:08, 19.84it/s]
 78%|#######8  | 610/781 [00:33<00:08, 19.85it/s]
 78%|#######8  | 612/781 [00:33<00:08, 19.85it/s]
 79%|#######8  | 614/781 [00:33<00:08, 19.83it/s]
 79%|#######8  | 616/781 [00:33<00:08, 19.83it/s]
 79%|#######9  | 618/781 [00:33<00:08, 19.85it/s]
 79%|#######9  | 620/781 [00:33<00:08, 19.70it/s]
 80%|#######9  | 622/781 [00:33<00:08, 19.73it/s]
 80%|#######9  | 624/781 [00:33<00:07, 19.73it/s]
 80%|########  | 626/781 [00:33<00:07, 19.73it/s]
 80%|########  | 628/781 [00:34<00:07, 19.37it/s]
 81%|########  | 630/781 [00:34<00:07, 19.37it/s]
 81%|########  | 632/781 [00:34<00:07, 19.36it/s]
 81%|########1 | 634/781 [00:34<00:07, 19.34it/s]
 81%|########1 | 636/781 [00:34<00:07, 19.30it/s]
 82%|########1 | 638/781 [00:34<00:07, 19.38it/s]
 82%|########1 | 640/781 [00:34<00:07, 19.50it/s]
 82%|########2 | 642/781 [00:34<00:07, 19.59it/s]
 82%|########2 | 644/781 [00:34<00:06, 19.66it/s]
 83%|########2 | 646/781 [00:35<00:06, 19.72it/s]
 83%|########2 | 648/781 [00:35<00:06, 19.46it/s]
 83%|########3 | 650/781 [00:35<00:06, 19.57it/s]
 83%|########3 | 652/781 [00:35<00:06, 19.66it/s]
 84%|########3 | 654/781 [00:35<00:06, 19.69it/s]
 84%|########3 | 656/781 [00:35<00:06, 19.73it/s]
 84%|########4 | 658/781 [00:35<00:06, 19.33it/s]
 85%|########4 | 660/781 [00:35<00:06, 19.05it/s]
 85%|########4 | 662/781 [00:35<00:06, 18.60it/s]
 85%|########5 | 664/781 [00:35<00:06, 18.29it/s]
 85%|########5 | 666/781 [00:36<00:06, 18.58it/s]
 86%|########5 | 668/781 [00:36<00:06, 18.59it/s]
 86%|########5 | 670/781 [00:36<00:05, 18.67it/s]
 86%|########6 | 672/781 [00:36<00:05, 18.48it/s]
 86%|########6 | 674/781 [00:36<00:05, 18.57it/s]
 87%|########6 | 676/781 [00:36<00:05, 18.43it/s]
 87%|########6 | 678/781 [00:36<00:05, 18.36it/s]
 87%|########7 | 680/781 [00:36<00:05, 18.35it/s]
 87%|########7 | 682/781 [00:36<00:05, 18.31it/s]
 88%|########7 | 684/781 [00:37<00:05, 18.71it/s]
 88%|########7 | 686/781 [00:37<00:05, 18.53it/s]
 88%|########8 | 688/781 [00:37<00:05, 18.40it/s]
 88%|########8 | 690/781 [00:37<00:04, 18.32it/s]
 89%|########8 | 692/781 [00:37<00:04, 18.11it/s]
 89%|########8 | 694/781 [00:37<00:04, 18.21it/s]
 89%|########9 | 696/781 [00:37<00:04, 18.37it/s]
 89%|########9 | 698/781 [00:37<00:04, 18.67it/s]
 90%|########9 | 700/781 [00:37<00:04, 18.30it/s]
 90%|########9 | 702/781 [00:38<00:04, 18.31it/s]
 90%|######### | 704/781 [00:38<00:04, 18.56it/s]
 90%|######### | 706/781 [00:38<00:04, 18.48it/s]
 91%|######### | 708/781 [00:38<00:03, 18.46it/s]
 91%|######### | 710/781 [00:38<00:03, 18.16it/s]
 91%|#########1| 712/781 [00:38<00:03, 18.52it/s]
 91%|#########1| 714/781 [00:38<00:03, 18.87it/s]
 92%|#########1| 716/781 [00:38<00:03, 19.16it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.09it/s]
 92%|#########2| 720/781 [00:38<00:03, 19.07it/s]
 92%|#########2| 722/781 [00:39<00:03, 19.06it/s]
 93%|#########2| 724/781 [00:39<00:03, 18.98it/s]
 93%|#########2| 726/781 [00:39<00:02, 18.85it/s]
 93%|#########3| 728/781 [00:39<00:02, 18.82it/s]
 93%|#########3| 730/781 [00:39<00:02, 18.78it/s]
 94%|#########3| 732/781 [00:39<00:02, 18.83it/s]
 94%|#########3| 734/781 [00:39<00:02, 18.60it/s]
 94%|#########4| 736/781 [00:39<00:02, 18.47it/s]
 94%|#########4| 738/781 [00:39<00:02, 18.41it/s]
 95%|#########4| 740/781 [00:40<00:02, 18.40it/s]
 95%|#########5| 742/781 [00:40<00:02, 18.36it/s]
 95%|#########5| 744/781 [00:40<00:02, 18.36it/s]
 96%|#########5| 746/781 [00:40<00:01, 18.03it/s]
 96%|#########5| 748/781 [00:40<00:01, 18.37it/s]
 96%|#########6| 750/781 [00:40<00:01, 18.51it/s]
 96%|#########6| 752/781 [00:40<00:01, 18.38it/s]
 97%|#########6| 754/781 [00:40<00:01, 18.27it/s]
 97%|#########6| 756/781 [00:40<00:01, 18.17it/s]
 97%|#########7| 758/781 [00:41<00:01, 18.07it/s]
 97%|#########7| 760/781 [00:41<00:01, 18.00it/s]
 98%|#########7| 762/781 [00:41<00:01, 18.20it/s]
 98%|#########7| 764/781 [00:41<00:00, 17.97it/s]
 98%|#########8| 766/781 [00:41<00:00, 17.81it/s]
 98%|#########8| 768/781 [00:41<00:00, 17.68it/s]
 99%|#########8| 770/781 [00:41<00:00, 17.59it/s]
 99%|#########8| 772/781 [00:41<00:00, 17.55it/s]
 99%|#########9| 774/781 [00:41<00:00, 17.75it/s]
 99%|#########9| 776/781 [00:42<00:00, 18.03it/s]
100%|#########9| 779/781 [00:42<00:00, 18.57it/s]
100%|##########| 781/781 [00:42<00:00, 18.46it/s]








Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.17s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:50,  1.20s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.26s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.18s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.18s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.18s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.18s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.18s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.18s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:46,  1.20s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.20s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.34s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.32s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:08,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.36s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.36s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.36s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.10s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.71s/it]
100%|##########| 1/1 [00:03<00:00,  3.71s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.13it/s]
  3%|2         | 4/157 [00:00<00:11, 13.53it/s]
  4%|3         | 6/157 [00:00<00:10, 13.93it/s]
  5%|5         | 8/157 [00:00<00:10, 14.36it/s]
  6%|6         | 10/157 [00:00<00:10, 14.70it/s]
  8%|7         | 12/157 [00:00<00:09, 14.87it/s]
  9%|8         | 14/157 [00:00<00:09, 14.89it/s]
 10%|#         | 16/157 [00:01<00:09, 14.92it/s]
 11%|#1        | 18/157 [00:01<00:09, 15.24it/s]
 13%|#2        | 20/157 [00:01<00:08, 15.47it/s]
 14%|#4        | 22/157 [00:01<00:08, 15.60it/s]
 15%|#5        | 24/157 [00:01<00:08, 15.55it/s]
 17%|#6        | 26/157 [00:01<00:08, 15.48it/s]
 18%|#7        | 28/157 [00:01<00:08, 15.79it/s]
 19%|#9        | 30/157 [00:01<00:08, 15.62it/s]
 20%|##        | 32/157 [00:02<00:07, 15.92it/s]
 22%|##1       | 34/157 [00:02<00:07, 15.87it/s]
 23%|##2       | 36/157 [00:02<00:07, 15.87it/s]
 24%|##4       | 38/157 [00:02<00:07, 15.60it/s]
 25%|##5       | 40/157 [00:02<00:07, 15.89it/s]
 27%|##6       | 42/157 [00:02<00:07, 16.12it/s]
 28%|##8       | 44/157 [00:02<00:06, 16.23it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.34it/s]
 31%|###       | 48/157 [00:03<00:06, 16.26it/s]
 32%|###1      | 50/157 [00:03<00:06, 16.28it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.31it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.31it/s]
 36%|###5      | 56/157 [00:03<00:06, 16.30it/s]
 37%|###6      | 58/157 [00:03<00:06, 16.33it/s]
 38%|###8      | 60/157 [00:03<00:05, 16.36it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.41it/s]
 41%|####      | 64/157 [00:04<00:05, 16.21it/s]
 42%|####2     | 66/157 [00:04<00:05, 16.06it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.08it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.14it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.16it/s]
 47%|####7     | 74/157 [00:04<00:05, 16.21it/s]
 48%|####8     | 76/157 [00:04<00:05, 15.99it/s]
 50%|####9     | 78/157 [00:04<00:04, 16.19it/s]
 51%|#####     | 80/157 [00:05<00:04, 16.38it/s]
 52%|#####2    | 82/157 [00:05<00:04, 16.48it/s]
 54%|#####3    | 84/157 [00:05<00:04, 16.63it/s]
 55%|#####4    | 86/157 [00:05<00:04, 16.52it/s]
 56%|#####6    | 88/157 [00:05<00:04, 16.52it/s]
 57%|#####7    | 90/157 [00:05<00:04, 16.48it/s]
 59%|#####8    | 92/157 [00:05<00:03, 16.37it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.30it/s]
 61%|######1   | 96/157 [00:06<00:03, 16.53it/s]
 62%|######2   | 98/157 [00:06<00:03, 16.59it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.49it/s]
 65%|######4   | 102/157 [00:06<00:03, 16.60it/s]
 66%|######6   | 104/157 [00:06<00:03, 16.58it/s]
 68%|######7   | 106/157 [00:06<00:03, 16.25it/s]
 69%|######8   | 108/157 [00:06<00:02, 16.44it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.54it/s]
 71%|#######1  | 112/157 [00:07<00:02, 16.17it/s]
 73%|#######2  | 114/157 [00:07<00:02, 16.27it/s]
 74%|#######3  | 116/157 [00:07<00:02, 16.35it/s]
 75%|#######5  | 118/157 [00:07<00:02, 16.38it/s]
 76%|#######6  | 120/157 [00:07<00:02, 16.33it/s]
 78%|#######7  | 122/157 [00:07<00:02, 16.25it/s]
 79%|#######8  | 124/157 [00:07<00:02, 16.23it/s]
 80%|########  | 126/157 [00:07<00:01, 16.40it/s]
 82%|########1 | 128/157 [00:07<00:01, 16.50it/s]
 83%|########2 | 130/157 [00:08<00:01, 16.47it/s]
 84%|########4 | 132/157 [00:08<00:01, 16.43it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.56it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.38it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.49it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.52it/s]
 90%|######### | 142/157 [00:08<00:00, 16.58it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.43it/s]
 93%|#########2| 146/157 [00:09<00:00, 16.35it/s]
 94%|#########4| 148/157 [00:09<00:00, 16.43it/s]
 96%|#########5| 150/157 [00:09<00:00, 16.49it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.51it/s]
 98%|#########8| 154/157 [00:09<00:00, 16.59it/s]
 99%|#########9| 156/157 [00:09<00:00, 16.68it/s]
100%|##########| 157/157 [00:09<00:00, 16.17it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:46, 16.88it/s]
  1%|          | 4/781 [00:00<00:45, 16.90it/s]
  1%|          | 6/781 [00:00<00:45, 17.15it/s]
  1%|1         | 8/781 [00:00<00:43, 17.86it/s]
  1%|1         | 10/781 [00:00<00:42, 18.05it/s]
  2%|1         | 12/781 [00:00<00:41, 18.38it/s]
  2%|1         | 14/781 [00:00<00:41, 18.39it/s]
  2%|2         | 16/781 [00:00<00:41, 18.50it/s]
  2%|2         | 18/781 [00:00<00:41, 18.59it/s]
  3%|2         | 20/781 [00:01<00:40, 18.73it/s]
  3%|2         | 22/781 [00:01<00:41, 18.49it/s]
  3%|3         | 24/781 [00:01<00:40, 18.61it/s]
  3%|3         | 26/781 [00:01<00:40, 18.72it/s]
  4%|3         | 28/781 [00:01<00:40, 18.77it/s]
  4%|3         | 30/781 [00:01<00:40, 18.77it/s]
  4%|4         | 32/781 [00:01<00:39, 18.85it/s]
  4%|4         | 34/781 [00:01<00:40, 18.66it/s]
  5%|4         | 36/781 [00:01<00:40, 18.58it/s]
  5%|4         | 38/781 [00:02<00:39, 18.62it/s]
  5%|5         | 40/781 [00:02<00:39, 18.73it/s]
  5%|5         | 42/781 [00:02<00:39, 18.61it/s]
  6%|5         | 44/781 [00:02<00:39, 18.54it/s]
  6%|5         | 46/781 [00:02<00:39, 18.57it/s]
  6%|6         | 48/781 [00:02<00:39, 18.79it/s]
  6%|6         | 50/781 [00:02<00:38, 18.95it/s]
  7%|6         | 52/781 [00:02<00:38, 18.87it/s]
  7%|6         | 54/781 [00:02<00:38, 18.69it/s]
  7%|7         | 56/781 [00:03<00:39, 18.59it/s]
  7%|7         | 58/781 [00:03<00:38, 18.61it/s]
  8%|7         | 60/781 [00:03<00:38, 18.61it/s]
  8%|7         | 62/781 [00:03<00:38, 18.74it/s]
  8%|8         | 64/781 [00:03<00:37, 18.89it/s]
  8%|8         | 66/781 [00:03<00:37, 18.96it/s]
  9%|8         | 68/781 [00:03<00:37, 19.00it/s]
  9%|8         | 70/781 [00:03<00:37, 18.99it/s]
  9%|9         | 72/781 [00:03<00:37, 18.99it/s]
  9%|9         | 74/781 [00:03<00:37, 18.98it/s]
 10%|9         | 76/781 [00:04<00:37, 18.98it/s]
 10%|9         | 78/781 [00:04<00:37, 18.93it/s]
 10%|#         | 80/781 [00:04<00:36, 18.98it/s]
 10%|#         | 82/781 [00:04<00:36, 18.95it/s]
 11%|#         | 84/781 [00:04<00:36, 18.93it/s]
 11%|#1        | 86/781 [00:04<00:36, 18.92it/s]
 11%|#1        | 88/781 [00:04<00:36, 18.95it/s]
 12%|#1        | 90/781 [00:04<00:36, 18.77it/s]
 12%|#1        | 92/781 [00:04<00:36, 18.74it/s]
 12%|#2        | 94/781 [00:05<00:36, 18.80it/s]
 12%|#2        | 96/781 [00:05<00:36, 18.90it/s]
 13%|#2        | 98/781 [00:05<00:36, 18.97it/s]
 13%|#2        | 100/781 [00:05<00:35, 19.00it/s]
 13%|#3        | 102/781 [00:05<00:35, 18.91it/s]
 13%|#3        | 104/781 [00:05<00:35, 18.96it/s]
 14%|#3        | 106/781 [00:05<00:35, 18.96it/s]
 14%|#3        | 108/781 [00:05<00:35, 18.91it/s]
 14%|#4        | 110/781 [00:05<00:35, 18.93it/s]
 14%|#4        | 112/781 [00:05<00:35, 18.86it/s]
 15%|#4        | 114/781 [00:06<00:35, 18.96it/s]
 15%|#4        | 116/781 [00:06<00:34, 19.03it/s]
 15%|#5        | 118/781 [00:06<00:34, 19.09it/s]
 15%|#5        | 120/781 [00:06<00:34, 19.00it/s]
 16%|#5        | 122/781 [00:06<00:34, 19.09it/s]
 16%|#5        | 124/781 [00:06<00:34, 19.01it/s]
 16%|#6        | 126/781 [00:06<00:34, 19.01it/s]
 16%|#6        | 128/781 [00:06<00:34, 18.90it/s]
 17%|#6        | 130/781 [00:06<00:34, 18.88it/s]
 17%|#6        | 132/781 [00:07<00:34, 18.58it/s]
 17%|#7        | 134/781 [00:07<00:34, 18.74it/s]
 17%|#7        | 136/781 [00:07<00:34, 18.68it/s]
 18%|#7        | 138/781 [00:07<00:34, 18.65it/s]
 18%|#7        | 140/781 [00:07<00:34, 18.58it/s]
 18%|#8        | 142/781 [00:07<00:34, 18.77it/s]
 18%|#8        | 144/781 [00:07<00:33, 18.86it/s]
 19%|#8        | 146/781 [00:07<00:33, 18.92it/s]
 19%|#8        | 148/781 [00:07<00:33, 18.84it/s]
 19%|#9        | 150/781 [00:08<00:33, 18.70it/s]
 19%|#9        | 152/781 [00:08<00:34, 18.50it/s]
 20%|#9        | 154/781 [00:08<00:33, 18.45it/s]
 20%|#9        | 156/781 [00:08<00:33, 18.60it/s]
 20%|##        | 158/781 [00:08<00:33, 18.73it/s]
 20%|##        | 160/781 [00:08<00:33, 18.77it/s]
 21%|##        | 162/781 [00:08<00:32, 18.86it/s]
 21%|##        | 164/781 [00:08<00:32, 18.95it/s]
 21%|##1       | 166/781 [00:08<00:32, 18.85it/s]
 22%|##1       | 168/781 [00:08<00:32, 18.99it/s]
 22%|##1       | 170/781 [00:09<00:32, 19.03it/s]
 22%|##2       | 172/781 [00:09<00:32, 18.90it/s]
 22%|##2       | 174/781 [00:09<00:32, 18.87it/s]
 23%|##2       | 176/781 [00:09<00:31, 18.95it/s]
 23%|##2       | 178/781 [00:09<00:32, 18.83it/s]
 23%|##3       | 180/781 [00:09<00:32, 18.75it/s]
 23%|##3       | 182/781 [00:09<00:31, 18.84it/s]
 24%|##3       | 184/781 [00:09<00:31, 18.94it/s]
 24%|##3       | 186/781 [00:09<00:31, 19.05it/s]
 24%|##4       | 188/781 [00:10<00:31, 19.12it/s]
 24%|##4       | 190/781 [00:10<00:30, 19.14it/s]
 25%|##4       | 192/781 [00:10<00:30, 19.14it/s]
 25%|##4       | 194/781 [00:10<00:30, 19.16it/s]
 25%|##5       | 196/781 [00:10<00:30, 19.21it/s]
 25%|##5       | 198/781 [00:10<00:30, 19.18it/s]
 26%|##5       | 200/781 [00:10<00:30, 18.96it/s]
 26%|##5       | 202/781 [00:10<00:30, 19.06it/s]
 26%|##6       | 204/781 [00:10<00:30, 19.04it/s]
 26%|##6       | 206/781 [00:10<00:30, 18.86it/s]
 27%|##6       | 208/781 [00:11<00:30, 18.86it/s]
 27%|##6       | 210/781 [00:11<00:30, 18.90it/s]
 27%|##7       | 212/781 [00:11<00:30, 18.79it/s]
 27%|##7       | 214/781 [00:11<00:30, 18.65it/s]
 28%|##7       | 216/781 [00:11<00:30, 18.52it/s]
 28%|##7       | 218/781 [00:11<00:30, 18.66it/s]
 28%|##8       | 220/781 [00:11<00:30, 18.65it/s]
 28%|##8       | 222/781 [00:11<00:30, 18.43it/s]
 29%|##8       | 224/781 [00:11<00:30, 18.52it/s]
 29%|##8       | 226/781 [00:12<00:30, 18.36it/s]
 29%|##9       | 228/781 [00:12<00:30, 18.24it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.40it/s]
 30%|##9       | 232/781 [00:12<00:29, 18.55it/s]
 30%|##9       | 234/781 [00:12<00:29, 18.31it/s]
 30%|###       | 236/781 [00:12<00:30, 18.11it/s]
 30%|###       | 238/781 [00:12<00:30, 18.08it/s]
 31%|###       | 240/781 [00:12<00:30, 17.75it/s]
 31%|###       | 242/781 [00:12<00:29, 17.98it/s]
 31%|###1      | 244/781 [00:13<00:29, 18.18it/s]
 31%|###1      | 246/781 [00:13<00:29, 18.11it/s]
 32%|###1      | 248/781 [00:13<00:28, 18.42it/s]
 32%|###2      | 250/781 [00:13<00:28, 18.68it/s]
 32%|###2      | 252/781 [00:13<00:28, 18.85it/s]
 33%|###2      | 254/781 [00:13<00:27, 18.87it/s]
 33%|###2      | 256/781 [00:13<00:27, 19.05it/s]
 33%|###3      | 258/781 [00:13<00:27, 19.06it/s]
 33%|###3      | 260/781 [00:13<00:27, 19.19it/s]
 34%|###3      | 262/781 [00:13<00:27, 19.16it/s]
 34%|###3      | 264/781 [00:14<00:27, 19.04it/s]
 34%|###4      | 266/781 [00:14<00:26, 19.16it/s]
 34%|###4      | 268/781 [00:14<00:26, 19.09it/s]
 35%|###4      | 270/781 [00:14<00:26, 19.16it/s]
 35%|###4      | 272/781 [00:14<00:26, 19.21it/s]
 35%|###5      | 274/781 [00:14<00:26, 19.24it/s]
 35%|###5      | 276/781 [00:14<00:26, 19.29it/s]
 36%|###5      | 278/781 [00:14<00:26, 19.29it/s]
 36%|###5      | 280/781 [00:14<00:26, 18.68it/s]
 36%|###6      | 282/781 [00:15<00:26, 18.49it/s]
 36%|###6      | 284/781 [00:15<00:27, 18.38it/s]
 37%|###6      | 286/781 [00:15<00:27, 18.15it/s]
 37%|###6      | 288/781 [00:15<00:27, 18.09it/s]
 37%|###7      | 290/781 [00:15<00:26, 18.23it/s]
 37%|###7      | 292/781 [00:15<00:26, 18.34it/s]
 38%|###7      | 294/781 [00:15<00:26, 18.29it/s]
 38%|###7      | 296/781 [00:15<00:26, 18.25it/s]
 38%|###8      | 298/781 [00:15<00:26, 18.19it/s]
 38%|###8      | 300/781 [00:16<00:26, 18.19it/s]
 39%|###8      | 302/781 [00:16<00:26, 18.17it/s]
 39%|###8      | 304/781 [00:16<00:26, 17.98it/s]
 39%|###9      | 306/781 [00:16<00:26, 17.86it/s]
 39%|###9      | 308/781 [00:16<00:26, 17.75it/s]
 40%|###9      | 310/781 [00:16<00:26, 17.74it/s]
 40%|###9      | 312/781 [00:16<00:26, 17.88it/s]
 40%|####      | 314/781 [00:16<00:25, 17.99it/s]
 40%|####      | 316/781 [00:16<00:25, 17.91it/s]
 41%|####      | 318/781 [00:17<00:25, 17.91it/s]
 41%|####      | 320/781 [00:17<00:25, 17.85it/s]
 41%|####1     | 322/781 [00:17<00:25, 17.88it/s]
 41%|####1     | 324/781 [00:17<00:25, 17.95it/s]
 42%|####1     | 326/781 [00:17<00:25, 18.01it/s]
 42%|####1     | 328/781 [00:17<00:24, 18.36it/s]
 42%|####2     | 330/781 [00:17<00:24, 18.34it/s]
 43%|####2     | 332/781 [00:17<00:24, 18.32it/s]
 43%|####2     | 334/781 [00:17<00:24, 18.49it/s]
 43%|####3     | 336/781 [00:18<00:23, 18.55it/s]
 43%|####3     | 338/781 [00:18<00:23, 18.59it/s]
 44%|####3     | 340/781 [00:18<00:23, 18.63it/s]
 44%|####3     | 342/781 [00:18<00:23, 18.71it/s]
 44%|####4     | 344/781 [00:18<00:23, 18.62it/s]
 44%|####4     | 346/781 [00:18<00:23, 18.61it/s]
 45%|####4     | 348/781 [00:18<00:23, 18.70it/s]
 45%|####4     | 350/781 [00:18<00:22, 18.80it/s]
 45%|####5     | 352/781 [00:18<00:22, 18.97it/s]
 45%|####5     | 354/781 [00:18<00:22, 19.05it/s]
 46%|####5     | 356/781 [00:19<00:22, 19.19it/s]
 46%|####5     | 358/781 [00:19<00:21, 19.31it/s]
 46%|####6     | 360/781 [00:19<00:21, 19.22it/s]
 46%|####6     | 362/781 [00:19<00:21, 19.23it/s]
 47%|####6     | 364/781 [00:19<00:21, 19.32it/s]
 47%|####6     | 366/781 [00:19<00:21, 19.39it/s]
 47%|####7     | 368/781 [00:19<00:21, 19.37it/s]
 47%|####7     | 370/781 [00:19<00:21, 19.21it/s]
 48%|####7     | 372/781 [00:19<00:22, 18.44it/s]
 48%|####7     | 374/781 [00:20<00:21, 18.64it/s]
 48%|####8     | 376/781 [00:20<00:21, 18.82it/s]
 48%|####8     | 378/781 [00:20<00:21, 19.02it/s]
 49%|####8     | 380/781 [00:20<00:20, 19.19it/s]
 49%|####8     | 382/781 [00:20<00:20, 19.37it/s]
 49%|####9     | 384/781 [00:20<00:20, 19.46it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.40it/s]
 50%|####9     | 388/781 [00:20<00:20, 19.30it/s]
 50%|####9     | 390/781 [00:20<00:20, 19.28it/s]
 50%|#####     | 392/781 [00:20<00:20, 18.54it/s]
 50%|#####     | 394/781 [00:21<00:20, 18.86it/s]
 51%|#####     | 396/781 [00:21<00:20, 18.93it/s]
 51%|#####     | 398/781 [00:21<00:20, 18.88it/s]
 51%|#####1    | 400/781 [00:21<00:20, 18.93it/s]
 51%|#####1    | 402/781 [00:21<00:19, 19.09it/s]
 52%|#####1    | 404/781 [00:21<00:19, 19.07it/s]
 52%|#####1    | 406/781 [00:21<00:19, 19.15it/s]
 52%|#####2    | 408/781 [00:21<00:19, 19.06it/s]
 52%|#####2    | 410/781 [00:21<00:19, 18.84it/s]
 53%|#####2    | 412/781 [00:22<00:19, 18.83it/s]
 53%|#####3    | 414/781 [00:22<00:19, 18.77it/s]
 53%|#####3    | 416/781 [00:22<00:19, 18.79it/s]
 54%|#####3    | 418/781 [00:22<00:19, 18.86it/s]
 54%|#####3    | 420/781 [00:22<00:18, 19.04it/s]
 54%|#####4    | 422/781 [00:22<00:18, 19.16it/s]
 54%|#####4    | 424/781 [00:22<00:18, 19.30it/s]
 55%|#####4    | 426/781 [00:22<00:18, 19.23it/s]
 55%|#####4    | 428/781 [00:22<00:18, 19.22it/s]
 55%|#####5    | 430/781 [00:22<00:18, 18.87it/s]
 55%|#####5    | 432/781 [00:23<00:18, 19.01it/s]
 56%|#####5    | 434/781 [00:23<00:18, 18.83it/s]
 56%|#####5    | 436/781 [00:23<00:18, 19.03it/s]
 56%|#####6    | 438/781 [00:23<00:18, 18.98it/s]
 56%|#####6    | 440/781 [00:23<00:17, 19.00it/s]
 57%|#####6    | 442/781 [00:23<00:17, 19.10it/s]
 57%|#####6    | 444/781 [00:23<00:17, 19.22it/s]
 57%|#####7    | 446/781 [00:23<00:17, 19.23it/s]
 57%|#####7    | 448/781 [00:23<00:17, 19.14it/s]
 58%|#####7    | 450/781 [00:24<00:17, 18.97it/s]
 58%|#####7    | 452/781 [00:24<00:17, 19.09it/s]
 58%|#####8    | 454/781 [00:24<00:17, 19.09it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.79it/s]
 59%|#####8    | 458/781 [00:24<00:17, 18.91it/s]
 59%|#####8    | 460/781 [00:24<00:17, 18.51it/s]
 59%|#####9    | 462/781 [00:24<00:17, 18.66it/s]
 59%|#####9    | 464/781 [00:24<00:16, 18.91it/s]
 60%|#####9    | 466/781 [00:24<00:16, 19.03it/s]
 60%|#####9    | 468/781 [00:24<00:16, 19.19it/s]
 60%|######    | 470/781 [00:25<00:16, 19.33it/s]
 60%|######    | 472/781 [00:25<00:16, 19.20it/s]
 61%|######    | 474/781 [00:25<00:15, 19.35it/s]
 61%|######    | 476/781 [00:25<00:15, 19.48it/s]
 61%|######1   | 478/781 [00:25<00:15, 19.49it/s]
 61%|######1   | 480/781 [00:25<00:15, 19.41it/s]
 62%|######1   | 482/781 [00:25<00:15, 19.29it/s]
 62%|######1   | 484/781 [00:25<00:15, 19.27it/s]
 62%|######2   | 486/781 [00:25<00:15, 19.34it/s]
 62%|######2   | 488/781 [00:25<00:15, 18.97it/s]
 63%|######2   | 490/781 [00:26<00:15, 18.90it/s]
 63%|######2   | 492/781 [00:26<00:15, 19.06it/s]
 63%|######3   | 494/781 [00:26<00:14, 19.21it/s]
 64%|######3   | 496/781 [00:26<00:14, 19.37it/s]
 64%|######3   | 498/781 [00:26<00:14, 19.40it/s]
 64%|######4   | 500/781 [00:26<00:14, 19.29it/s]
 64%|######4   | 502/781 [00:26<00:14, 19.05it/s]
 65%|######4   | 504/781 [00:26<00:14, 18.94it/s]
 65%|######4   | 506/781 [00:27<00:19, 14.42it/s]
 65%|######5   | 508/781 [00:27<00:17, 15.63it/s]
 65%|######5   | 510/781 [00:27<00:16, 16.36it/s]
 66%|######5   | 512/781 [00:27<00:15, 17.03it/s]
 66%|######5   | 514/781 [00:27<00:15, 17.68it/s]
 66%|######6   | 516/781 [00:27<00:14, 18.04it/s]
 66%|######6   | 518/781 [00:27<00:14, 18.37it/s]
 67%|######6   | 520/781 [00:27<00:13, 18.72it/s]
 67%|######6   | 522/781 [00:27<00:13, 18.99it/s]
 67%|######7   | 524/781 [00:27<00:13, 18.91it/s]
 67%|######7   | 526/781 [00:28<00:13, 18.30it/s]
 68%|######7   | 528/781 [00:28<00:13, 18.39it/s]
 68%|######7   | 530/781 [00:28<00:13, 18.47it/s]
 68%|######8   | 532/781 [00:28<00:13, 18.57it/s]
 68%|######8   | 534/781 [00:28<00:13, 18.78it/s]
 69%|######8   | 536/781 [00:28<00:12, 19.03it/s]
 69%|######8   | 538/781 [00:28<00:12, 19.05it/s]
 69%|######9   | 540/781 [00:28<00:12, 19.21it/s]
 69%|######9   | 542/781 [00:28<00:12, 19.37it/s]
 70%|######9   | 544/781 [00:29<00:12, 19.43it/s]
 70%|######9   | 546/781 [00:29<00:12, 19.45it/s]
 70%|#######   | 548/781 [00:29<00:11, 19.43it/s]
 70%|#######   | 550/781 [00:29<00:11, 19.31it/s]
 71%|#######   | 552/781 [00:29<00:11, 19.24it/s]
 71%|#######   | 554/781 [00:29<00:11, 19.31it/s]
 71%|#######1  | 556/781 [00:29<00:11, 19.22it/s]
 71%|#######1  | 558/781 [00:29<00:11, 19.20it/s]
 72%|#######1  | 560/781 [00:29<00:11, 19.36it/s]
 72%|#######1  | 562/781 [00:29<00:11, 19.38it/s]
 72%|#######2  | 564/781 [00:30<00:11, 19.27it/s]
 72%|#######2  | 566/781 [00:30<00:11, 19.31it/s]
 73%|#######2  | 568/781 [00:30<00:11, 19.18it/s]
 73%|#######2  | 570/781 [00:30<00:11, 19.07it/s]
 73%|#######3  | 572/781 [00:30<00:10, 19.07it/s]
 73%|#######3  | 574/781 [00:30<00:10, 19.27it/s]
 74%|#######3  | 576/781 [00:30<00:10, 19.42it/s]
 74%|#######4  | 578/781 [00:30<00:10, 19.51it/s]
 74%|#######4  | 580/781 [00:30<00:10, 19.57it/s]
 75%|#######4  | 582/781 [00:31<00:10, 19.54it/s]
 75%|#######4  | 584/781 [00:31<00:10, 19.57it/s]
 75%|#######5  | 586/781 [00:31<00:10, 19.27it/s]
 75%|#######5  | 588/781 [00:31<00:10, 19.14it/s]
 76%|#######5  | 590/781 [00:31<00:10, 18.95it/s]
 76%|#######5  | 592/781 [00:31<00:09, 19.10it/s]
 76%|#######6  | 594/781 [00:31<00:09, 18.93it/s]
 76%|#######6  | 596/781 [00:31<00:09, 19.08it/s]
 77%|#######6  | 598/781 [00:31<00:09, 19.09it/s]
 77%|#######6  | 600/781 [00:31<00:09, 19.12it/s]
 77%|#######7  | 602/781 [00:32<00:09, 18.81it/s]
 77%|#######7  | 604/781 [00:32<00:09, 18.61it/s]
 78%|#######7  | 606/781 [00:32<00:09, 18.19it/s]
 78%|#######7  | 608/781 [00:32<00:09, 18.28it/s]
 78%|#######8  | 610/781 [00:32<00:09, 18.30it/s]
 78%|#######8  | 612/781 [00:32<00:09, 18.15it/s]
 79%|#######8  | 614/781 [00:32<00:09, 18.27it/s]
 79%|#######8  | 616/781 [00:32<00:08, 18.45it/s]
 79%|#######9  | 618/781 [00:32<00:08, 18.62it/s]
 79%|#######9  | 620/781 [00:33<00:08, 18.65it/s]
 80%|#######9  | 622/781 [00:33<00:08, 18.73it/s]
 80%|#######9  | 624/781 [00:33<00:08, 18.78it/s]
 80%|########  | 626/781 [00:33<00:08, 18.81it/s]
 80%|########  | 628/781 [00:33<00:08, 18.79it/s]
 81%|########  | 630/781 [00:33<00:07, 18.93it/s]
 81%|########  | 632/781 [00:33<00:07, 18.64it/s]
 81%|########1 | 634/781 [00:33<00:07, 18.81it/s]
 81%|########1 | 636/781 [00:33<00:07, 18.89it/s]
 82%|########1 | 638/781 [00:33<00:07, 18.90it/s]
 82%|########1 | 640/781 [00:34<00:07, 18.76it/s]
 82%|########2 | 642/781 [00:34<00:07, 18.89it/s]
 82%|########2 | 644/781 [00:34<00:07, 18.95it/s]
 83%|########2 | 646/781 [00:34<00:07, 18.69it/s]
 83%|########2 | 648/781 [00:34<00:07, 18.61it/s]
 83%|########3 | 650/781 [00:34<00:07, 18.52it/s]
 83%|########3 | 652/781 [00:34<00:06, 18.54it/s]
 84%|########3 | 654/781 [00:34<00:06, 18.55it/s]
 84%|########3 | 656/781 [00:34<00:06, 18.71it/s]
 84%|########4 | 658/781 [00:35<00:06, 18.85it/s]
 85%|########4 | 660/781 [00:35<00:06, 18.94it/s]
 85%|########4 | 662/781 [00:35<00:06, 18.98it/s]
 85%|########5 | 664/781 [00:35<00:06, 18.98it/s]
 85%|########5 | 666/781 [00:35<00:06, 18.93it/s]
 86%|########5 | 668/781 [00:35<00:06, 18.65it/s]
 86%|########5 | 670/781 [00:35<00:05, 18.74it/s]
 86%|########6 | 672/781 [00:35<00:05, 18.61it/s]
 86%|########6 | 674/781 [00:35<00:05, 18.70it/s]
 87%|########6 | 676/781 [00:36<00:05, 18.83it/s]
 87%|########6 | 678/781 [00:36<00:05, 18.90it/s]
 87%|########7 | 680/781 [00:36<00:05, 19.01it/s]
 87%|########7 | 682/781 [00:36<00:05, 19.05it/s]
 88%|########7 | 684/781 [00:36<00:05, 18.92it/s]
 88%|########7 | 686/781 [00:36<00:05, 18.83it/s]
 88%|########8 | 688/781 [00:36<00:04, 18.84it/s]
 88%|########8 | 690/781 [00:36<00:04, 18.90it/s]
 89%|########8 | 692/781 [00:36<00:04, 18.96it/s]
 89%|########8 | 694/781 [00:36<00:04, 19.05it/s]
 89%|########9 | 696/781 [00:37<00:04, 19.12it/s]
 89%|########9 | 698/781 [00:37<00:04, 19.09it/s]
 90%|########9 | 700/781 [00:37<00:04, 19.12it/s]
 90%|########9 | 702/781 [00:37<00:04, 18.75it/s]
 90%|######### | 704/781 [00:37<00:04, 18.69it/s]
 90%|######### | 706/781 [00:37<00:04, 18.75it/s]
 91%|######### | 708/781 [00:37<00:03, 18.87it/s]
 91%|######### | 710/781 [00:37<00:03, 18.93it/s]
 91%|#########1| 712/781 [00:37<00:03, 18.82it/s]
 91%|#########1| 714/781 [00:38<00:03, 18.94it/s]
 92%|#########1| 716/781 [00:38<00:03, 19.02it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.05it/s]
 92%|#########2| 720/781 [00:38<00:03, 18.85it/s]
 92%|#########2| 722/781 [00:38<00:03, 18.87it/s]
 93%|#########2| 724/781 [00:38<00:03, 18.95it/s]
 93%|#########2| 726/781 [00:38<00:02, 18.99it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.07it/s]
 93%|#########3| 730/781 [00:38<00:02, 19.06it/s]
 94%|#########3| 732/781 [00:38<00:02, 19.05it/s]
 94%|#########3| 734/781 [00:39<00:02, 19.05it/s]
 94%|#########4| 736/781 [00:39<00:02, 19.06it/s]
 94%|#########4| 738/781 [00:39<00:02, 19.02it/s]
 95%|#########4| 740/781 [00:39<00:02, 19.06it/s]
 95%|#########5| 742/781 [00:39<00:02, 18.81it/s]
 95%|#########5| 744/781 [00:39<00:01, 18.93it/s]
 96%|#########5| 746/781 [00:39<00:01, 18.72it/s]
 96%|#########5| 748/781 [00:39<00:01, 18.74it/s]
 96%|#########6| 750/781 [00:39<00:01, 18.72it/s]
 96%|#########6| 752/781 [00:40<00:01, 18.88it/s]
 97%|#########6| 754/781 [00:40<00:01, 18.88it/s]
 97%|#########6| 756/781 [00:40<00:01, 18.91it/s]
 97%|#########7| 758/781 [00:40<00:01, 19.01it/s]
 97%|#########7| 760/781 [00:40<00:01, 18.87it/s]
 98%|#########7| 762/781 [00:40<00:01, 18.87it/s]
 98%|#########7| 764/781 [00:40<00:00, 18.89it/s]
 98%|#########8| 766/781 [00:40<00:00, 18.99it/s]
 98%|#########8| 768/781 [00:40<00:00, 19.09it/s]
 99%|#########8| 770/781 [00:40<00:00, 18.93it/s]
 99%|#########8| 772/781 [00:41<00:00, 18.98it/s]
 99%|#########9| 774/781 [00:41<00:00, 18.99it/s]
 99%|#########9| 776/781 [00:41<00:00, 18.94it/s]
100%|#########9| 778/781 [00:41<00:00, 18.90it/s]
100%|#########9| 780/781 [00:41<00:00, 18.85it/s]
100%|##########| 781/781 [00:41<00:00, 18.79it/s]








Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.26s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:46,  1.18s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:41,  1.19s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.21s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.28s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.28s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.28s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.26s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:04<00:00,  4.21s/it]
100%|##########| 1/1 [00:04<00:00,  4.21s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.35it/s]
  3%|2         | 4/157 [00:00<00:11, 13.49it/s]
  4%|3         | 6/157 [00:00<00:11, 13.61it/s]
  5%|5         | 8/157 [00:00<00:10, 14.57it/s]
  6%|6         | 10/157 [00:00<00:09, 15.16it/s]
  8%|7         | 12/157 [00:00<00:09, 15.40it/s]
  9%|8         | 14/157 [00:00<00:09, 15.70it/s]
 10%|#         | 16/157 [00:01<00:08, 15.89it/s]
 11%|#1        | 18/157 [00:01<00:08, 15.91it/s]
 13%|#2        | 20/157 [00:01<00:08, 16.15it/s]
 14%|#4        | 22/157 [00:01<00:08, 16.36it/s]
 15%|#5        | 24/157 [00:01<00:08, 16.38it/s]
 17%|#6        | 26/157 [00:01<00:07, 16.56it/s]
 18%|#7        | 28/157 [00:01<00:07, 16.59it/s]
 19%|#9        | 30/157 [00:01<00:07, 16.56it/s]
 20%|##        | 32/157 [00:02<00:07, 16.62it/s]
 22%|##1       | 34/157 [00:02<00:07, 16.77it/s]
 23%|##2       | 36/157 [00:02<00:07, 16.85it/s]
 24%|##4       | 38/157 [00:02<00:07, 16.77it/s]
 25%|##5       | 40/157 [00:02<00:06, 16.83it/s]
 27%|##6       | 42/157 [00:02<00:06, 16.84it/s]
 28%|##8       | 44/157 [00:02<00:06, 16.83it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.85it/s]
 31%|###       | 48/157 [00:02<00:06, 16.89it/s]
 32%|###1      | 50/157 [00:03<00:06, 16.88it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.87it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.67it/s]
 36%|###5      | 56/157 [00:03<00:06, 16.71it/s]
 37%|###6      | 58/157 [00:03<00:05, 16.78it/s]
 38%|###8      | 60/157 [00:03<00:05, 16.65it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.46it/s]
 41%|####      | 64/157 [00:03<00:05, 16.21it/s]
 42%|####2     | 66/157 [00:04<00:05, 16.42it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.54it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.45it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.54it/s]
 47%|####7     | 74/157 [00:04<00:04, 16.66it/s]
 48%|####8     | 76/157 [00:04<00:04, 16.73it/s]
 50%|####9     | 78/157 [00:04<00:04, 16.39it/s]
 51%|#####     | 80/157 [00:04<00:04, 16.53it/s]
 52%|#####2    | 82/157 [00:05<00:04, 16.55it/s]
 54%|#####3    | 84/157 [00:05<00:04, 16.64it/s]
 55%|#####4    | 86/157 [00:05<00:04, 16.50it/s]
 56%|#####6    | 88/157 [00:05<00:04, 16.34it/s]
 57%|#####7    | 90/157 [00:05<00:04, 16.28it/s]
 59%|#####8    | 92/157 [00:05<00:03, 16.32it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.38it/s]
 61%|######1   | 96/157 [00:05<00:03, 16.36it/s]
 62%|######2   | 98/157 [00:06<00:03, 16.35it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.49it/s]
 65%|######4   | 102/157 [00:06<00:03, 16.52it/s]
 66%|######6   | 104/157 [00:06<00:03, 16.57it/s]
 68%|######7   | 106/157 [00:06<00:03, 16.65it/s]
 69%|######8   | 108/157 [00:06<00:02, 16.74it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.76it/s]
 71%|#######1  | 112/157 [00:06<00:02, 16.81it/s]
 73%|#######2  | 114/157 [00:06<00:02, 16.83it/s]
 74%|#######3  | 116/157 [00:07<00:02, 16.85it/s]
 75%|#######5  | 118/157 [00:07<00:02, 16.88it/s]
 76%|#######6  | 120/157 [00:07<00:02, 16.88it/s]
 78%|#######7  | 122/157 [00:07<00:02, 16.93it/s]
 79%|#######8  | 124/157 [00:07<00:01, 16.88it/s]
 80%|########  | 126/157 [00:07<00:01, 16.79it/s]
 82%|########1 | 128/157 [00:07<00:01, 16.69it/s]
 83%|########2 | 130/157 [00:07<00:01, 16.70it/s]
 84%|########4 | 132/157 [00:08<00:01, 16.77it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.82it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.83it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.78it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.79it/s]
 90%|######### | 142/157 [00:08<00:00, 16.77it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.79it/s]
 93%|#########2| 146/157 [00:08<00:00, 16.73it/s]
 94%|#########4| 148/157 [00:08<00:00, 16.70it/s]
 96%|#########5| 150/157 [00:09<00:00, 16.61it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.63it/s]
 98%|#########8| 154/157 [00:09<00:00, 16.75it/s]
 99%|#########9| 156/157 [00:09<00:00, 16.72it/s]
100%|##########| 157/157 [00:09<00:00, 16.55it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:45, 17.25it/s]
  1%|          | 4/781 [00:00<00:45, 16.99it/s]
  1%|          | 6/781 [00:00<00:44, 17.57it/s]
  1%|1         | 8/781 [00:00<00:42, 18.04it/s]
  1%|1         | 10/781 [00:00<00:42, 18.25it/s]
  2%|1         | 12/781 [00:00<00:41, 18.52it/s]
  2%|1         | 14/781 [00:00<00:41, 18.70it/s]
  2%|2         | 16/781 [00:00<00:40, 18.82it/s]
  2%|2         | 18/781 [00:00<00:40, 18.93it/s]
  3%|2         | 20/781 [00:01<00:40, 18.96it/s]
  3%|2         | 22/781 [00:01<00:39, 18.99it/s]
  3%|3         | 24/781 [00:01<00:39, 18.97it/s]
  3%|3         | 26/781 [00:01<00:40, 18.61it/s]
  4%|3         | 28/781 [00:01<00:40, 18.39it/s]
  4%|3         | 30/781 [00:01<00:40, 18.35it/s]
  4%|4         | 32/781 [00:01<00:40, 18.30it/s]
  4%|4         | 34/781 [00:01<00:40, 18.30it/s]
  5%|4         | 36/781 [00:01<00:40, 18.53it/s]
  5%|4         | 38/781 [00:02<00:39, 18.68it/s]
  5%|5         | 40/781 [00:02<00:39, 18.77it/s]
  5%|5         | 42/781 [00:02<00:39, 18.83it/s]
  6%|5         | 44/781 [00:02<00:38, 18.93it/s]
  6%|5         | 46/781 [00:02<00:38, 18.87it/s]
  6%|6         | 48/781 [00:02<00:38, 18.88it/s]
  6%|6         | 50/781 [00:02<00:38, 18.96it/s]
  7%|6         | 52/781 [00:02<00:38, 19.03it/s]
  7%|6         | 54/781 [00:02<00:38, 18.85it/s]
  7%|7         | 56/781 [00:03<00:38, 18.70it/s]
  7%|7         | 58/781 [00:03<00:38, 18.65it/s]
  8%|7         | 60/781 [00:03<00:38, 18.73it/s]
  8%|7         | 62/781 [00:03<00:38, 18.87it/s]
  8%|8         | 64/781 [00:03<00:37, 18.89it/s]
  8%|8         | 66/781 [00:03<00:37, 18.97it/s]
  9%|8         | 68/781 [00:03<00:37, 18.94it/s]
  9%|8         | 70/781 [00:03<00:37, 19.00it/s]
  9%|9         | 72/781 [00:03<00:37, 19.05it/s]
  9%|9         | 74/781 [00:03<00:37, 19.10it/s]
 10%|9         | 76/781 [00:04<00:36, 19.09it/s]
 10%|9         | 78/781 [00:04<00:37, 18.85it/s]
 10%|#         | 80/781 [00:04<00:37, 18.73it/s]
 10%|#         | 82/781 [00:04<00:37, 18.78it/s]
 11%|#         | 84/781 [00:04<00:37, 18.76it/s]
 11%|#1        | 86/781 [00:04<00:36, 18.79it/s]
 11%|#1        | 88/781 [00:04<00:36, 18.84it/s]
 12%|#1        | 90/781 [00:04<00:36, 18.76it/s]
 12%|#1        | 92/781 [00:04<00:36, 18.85it/s]
 12%|#2        | 94/781 [00:05<00:36, 18.73it/s]
 12%|#2        | 96/781 [00:05<00:36, 18.76it/s]
 13%|#2        | 98/781 [00:05<00:36, 18.77it/s]
 13%|#2        | 100/781 [00:05<00:36, 18.75it/s]
 13%|#3        | 102/781 [00:05<00:36, 18.81it/s]
 13%|#3        | 104/781 [00:05<00:35, 18.90it/s]
 14%|#3        | 106/781 [00:05<00:35, 18.85it/s]
 14%|#3        | 108/781 [00:05<00:35, 18.82it/s]
 14%|#4        | 110/781 [00:05<00:35, 18.89it/s]
 14%|#4        | 112/781 [00:05<00:35, 18.90it/s]
 15%|#4        | 114/781 [00:06<00:35, 18.88it/s]
 15%|#4        | 116/781 [00:06<00:35, 18.82it/s]
 15%|#5        | 118/781 [00:06<00:35, 18.89it/s]
 15%|#5        | 120/781 [00:06<00:34, 18.92it/s]
 16%|#5        | 122/781 [00:06<00:34, 18.91it/s]
 16%|#5        | 124/781 [00:06<00:34, 18.91it/s]
 16%|#6        | 126/781 [00:06<00:34, 18.95it/s]
 16%|#6        | 128/781 [00:06<00:34, 18.87it/s]
 17%|#6        | 130/781 [00:06<00:34, 18.92it/s]
 17%|#6        | 132/781 [00:07<00:34, 18.94it/s]
 17%|#7        | 134/781 [00:07<00:34, 18.84it/s]
 17%|#7        | 136/781 [00:07<00:34, 18.64it/s]
 18%|#7        | 138/781 [00:07<00:34, 18.78it/s]
 18%|#7        | 140/781 [00:07<00:33, 18.90it/s]
 18%|#8        | 142/781 [00:07<00:33, 18.94it/s]
 18%|#8        | 144/781 [00:07<00:33, 18.88it/s]
 19%|#8        | 146/781 [00:07<00:33, 18.89it/s]
 19%|#8        | 148/781 [00:07<00:33, 18.87it/s]
 19%|#9        | 150/781 [00:07<00:33, 19.01it/s]
 19%|#9        | 152/781 [00:08<00:32, 19.10it/s]
 20%|#9        | 154/781 [00:08<00:32, 19.04it/s]
 20%|#9        | 156/781 [00:08<00:32, 18.95it/s]
 20%|##        | 158/781 [00:08<00:33, 18.82it/s]
 20%|##        | 160/781 [00:08<00:33, 18.80it/s]
 21%|##        | 162/781 [00:08<00:33, 18.74it/s]
 21%|##        | 164/781 [00:08<00:32, 18.90it/s]
 21%|##1       | 166/781 [00:08<00:32, 18.95it/s]
 22%|##1       | 168/781 [00:08<00:32, 18.83it/s]
 22%|##1       | 170/781 [00:09<00:32, 18.69it/s]
 22%|##2       | 172/781 [00:09<00:32, 18.65it/s]
 22%|##2       | 174/781 [00:09<00:32, 18.61it/s]
 23%|##2       | 176/781 [00:09<00:32, 18.70it/s]
 23%|##2       | 178/781 [00:09<00:32, 18.57it/s]
 23%|##3       | 180/781 [00:09<00:32, 18.65it/s]
 23%|##3       | 182/781 [00:09<00:32, 18.70it/s]
 24%|##3       | 184/781 [00:09<00:31, 18.69it/s]
 24%|##3       | 186/781 [00:09<00:31, 18.65it/s]
 24%|##4       | 188/781 [00:10<00:31, 18.72it/s]
 24%|##4       | 190/781 [00:10<00:31, 18.77it/s]
 25%|##4       | 192/781 [00:10<00:31, 18.71it/s]
 25%|##4       | 194/781 [00:10<00:31, 18.75it/s]
 25%|##5       | 196/781 [00:10<00:31, 18.59it/s]
 25%|##5       | 198/781 [00:10<00:31, 18.63it/s]
 26%|##5       | 200/781 [00:10<00:31, 18.66it/s]
 26%|##5       | 202/781 [00:10<00:30, 18.76it/s]
 26%|##6       | 204/781 [00:10<00:31, 18.58it/s]
 26%|##6       | 206/781 [00:10<00:31, 18.22it/s]
 27%|##6       | 208/781 [00:11<00:30, 18.57it/s]
 27%|##6       | 210/781 [00:11<00:30, 18.59it/s]
 27%|##7       | 212/781 [00:11<00:30, 18.71it/s]
 27%|##7       | 214/781 [00:11<00:30, 18.38it/s]
 28%|##7       | 216/781 [00:11<00:30, 18.43it/s]
 28%|##7       | 218/781 [00:11<00:30, 18.31it/s]
 28%|##8       | 220/781 [00:11<00:30, 18.42it/s]
 28%|##8       | 222/781 [00:11<00:29, 18.70it/s]
 29%|##8       | 224/781 [00:11<00:29, 18.90it/s]
 29%|##8       | 226/781 [00:12<00:30, 18.20it/s]
 29%|##9       | 228/781 [00:12<00:29, 18.56it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.80it/s]
 30%|##9       | 232/781 [00:12<00:28, 18.96it/s]
 30%|##9       | 234/781 [00:12<00:29, 18.71it/s]
 30%|###       | 236/781 [00:12<00:28, 18.84it/s]
 30%|###       | 238/781 [00:12<00:28, 18.95it/s]
 31%|###       | 240/781 [00:12<00:28, 19.05it/s]
 31%|###       | 242/781 [00:12<00:28, 19.15it/s]
 31%|###1      | 244/781 [00:13<00:27, 19.25it/s]
 31%|###1      | 246/781 [00:13<00:27, 19.29it/s]
 32%|###1      | 248/781 [00:13<00:27, 19.29it/s]
 32%|###2      | 250/781 [00:13<00:27, 19.10it/s]
 32%|###2      | 252/781 [00:13<00:27, 19.04it/s]
 33%|###2      | 254/781 [00:13<00:27, 18.94it/s]
 33%|###2      | 256/781 [00:13<00:27, 18.95it/s]
 33%|###3      | 258/781 [00:13<00:27, 18.95it/s]
 33%|###3      | 260/781 [00:13<00:27, 19.02it/s]
 34%|###3      | 262/781 [00:13<00:27, 19.13it/s]
 34%|###3      | 264/781 [00:14<00:27, 19.07it/s]
 34%|###4      | 266/781 [00:14<00:27, 18.91it/s]
 34%|###4      | 268/781 [00:14<00:27, 18.75it/s]
 35%|###4      | 270/781 [00:14<00:27, 18.37it/s]
 35%|###4      | 272/781 [00:14<00:27, 18.37it/s]
 35%|###5      | 274/781 [00:14<00:27, 18.23it/s]
 35%|###5      | 276/781 [00:14<00:27, 18.07it/s]
 36%|###5      | 278/781 [00:14<00:27, 18.07it/s]
 36%|###5      | 280/781 [00:14<00:27, 18.06it/s]
 36%|###6      | 282/781 [00:15<00:27, 18.28it/s]
 36%|###6      | 284/781 [00:15<00:27, 18.22it/s]
 37%|###6      | 286/781 [00:15<00:26, 18.47it/s]
 37%|###6      | 288/781 [00:15<00:26, 18.61it/s]
 37%|###7      | 290/781 [00:15<00:26, 18.61it/s]
 37%|###7      | 292/781 [00:15<00:26, 18.61it/s]
 38%|###7      | 294/781 [00:15<00:26, 18.72it/s]
 38%|###7      | 296/781 [00:15<00:25, 18.78it/s]
 38%|###8      | 298/781 [00:15<00:25, 18.73it/s]
 38%|###8      | 300/781 [00:16<00:25, 18.72it/s]
 39%|###8      | 302/781 [00:16<00:25, 18.71it/s]
 39%|###8      | 304/781 [00:16<00:25, 18.74it/s]
 39%|###9      | 306/781 [00:16<00:25, 18.86it/s]
 39%|###9      | 308/781 [00:16<00:24, 18.96it/s]
 40%|###9      | 310/781 [00:16<00:24, 18.98it/s]
 40%|###9      | 312/781 [00:16<00:24, 18.88it/s]
 40%|####      | 314/781 [00:16<00:24, 18.94it/s]
 40%|####      | 316/781 [00:16<00:24, 19.00it/s]
 41%|####      | 318/781 [00:16<00:24, 18.88it/s]
 41%|####      | 320/781 [00:17<00:24, 18.67it/s]
 41%|####1     | 322/781 [00:17<00:24, 18.81it/s]
 41%|####1     | 324/781 [00:17<00:24, 18.66it/s]
 42%|####1     | 326/781 [00:17<00:24, 18.72it/s]
 42%|####1     | 328/781 [00:17<00:24, 18.80it/s]
 42%|####2     | 330/781 [00:17<00:23, 18.85it/s]
 43%|####2     | 332/781 [00:17<00:23, 18.78it/s]
 43%|####2     | 334/781 [00:17<00:23, 18.85it/s]
 43%|####3     | 336/781 [00:17<00:23, 18.95it/s]
 43%|####3     | 338/781 [00:18<00:23, 19.05it/s]
 44%|####3     | 340/781 [00:18<00:23, 19.03it/s]
 44%|####3     | 342/781 [00:18<00:23, 19.05it/s]
 44%|####4     | 344/781 [00:18<00:23, 18.90it/s]
 44%|####4     | 346/781 [00:18<00:23, 18.80it/s]
 45%|####4     | 348/781 [00:18<00:23, 18.76it/s]
 45%|####4     | 350/781 [00:18<00:22, 18.81it/s]
 45%|####5     | 352/781 [00:18<00:22, 18.95it/s]
 45%|####5     | 354/781 [00:18<00:22, 18.92it/s]
 46%|####5     | 356/781 [00:18<00:22, 18.87it/s]
 46%|####5     | 358/781 [00:19<00:22, 18.76it/s]
 46%|####6     | 360/781 [00:19<00:22, 18.70it/s]
 46%|####6     | 362/781 [00:19<00:22, 18.57it/s]
 47%|####6     | 364/781 [00:19<00:22, 18.66it/s]
 47%|####6     | 366/781 [00:19<00:22, 18.76it/s]
 47%|####7     | 368/781 [00:19<00:22, 18.69it/s]
 47%|####7     | 370/781 [00:19<00:21, 18.70it/s]
 48%|####7     | 372/781 [00:19<00:21, 18.68it/s]
 48%|####7     | 374/781 [00:19<00:21, 18.74it/s]
 48%|####8     | 376/781 [00:20<00:21, 18.67it/s]
 48%|####8     | 378/781 [00:20<00:21, 18.74it/s]
 49%|####8     | 380/781 [00:20<00:21, 18.70it/s]
 49%|####8     | 382/781 [00:20<00:21, 18.75it/s]
 49%|####9     | 384/781 [00:20<00:21, 18.70it/s]
 49%|####9     | 386/781 [00:20<00:21, 18.79it/s]
 50%|####9     | 388/781 [00:20<00:20, 18.78it/s]
 50%|####9     | 390/781 [00:20<00:20, 18.67it/s]
 50%|#####     | 392/781 [00:20<00:20, 18.66it/s]
 50%|#####     | 394/781 [00:21<00:20, 18.79it/s]
 51%|#####     | 396/781 [00:21<00:20, 18.81it/s]
 51%|#####     | 398/781 [00:21<00:20, 18.87it/s]
 51%|#####1    | 400/781 [00:21<00:20, 18.93it/s]
 51%|#####1    | 402/781 [00:21<00:19, 18.96it/s]
 52%|#####1    | 404/781 [00:21<00:19, 18.97it/s]
 52%|#####1    | 406/781 [00:21<00:19, 18.90it/s]
 52%|#####2    | 408/781 [00:21<00:19, 18.83it/s]
 52%|#####2    | 410/781 [00:21<00:19, 18.87it/s]
 53%|#####2    | 412/781 [00:21<00:19, 18.90it/s]
 53%|#####3    | 414/781 [00:22<00:19, 18.82it/s]
 53%|#####3    | 416/781 [00:22<00:19, 18.79it/s]
 54%|#####3    | 418/781 [00:22<00:19, 18.90it/s]
 54%|#####3    | 420/781 [00:22<00:19, 18.93it/s]
 54%|#####4    | 422/781 [00:22<00:19, 18.78it/s]
 54%|#####4    | 424/781 [00:22<00:18, 18.79it/s]
 55%|#####4    | 426/781 [00:22<00:18, 18.89it/s]
 55%|#####4    | 428/781 [00:22<00:18, 18.69it/s]
 55%|#####5    | 430/781 [00:22<00:18, 18.78it/s]
 55%|#####5    | 432/781 [00:23<00:18, 18.71it/s]
 56%|#####5    | 434/781 [00:23<00:18, 18.76it/s]
 56%|#####5    | 436/781 [00:23<00:18, 18.81it/s]
 56%|#####6    | 438/781 [00:23<00:18, 18.77it/s]
 56%|#####6    | 440/781 [00:23<00:18, 18.69it/s]
 57%|#####6    | 442/781 [00:23<00:17, 18.86it/s]
 57%|#####6    | 444/781 [00:23<00:17, 18.98it/s]
 57%|#####7    | 446/781 [00:23<00:17, 19.02it/s]
 57%|#####7    | 448/781 [00:23<00:17, 19.01it/s]
 58%|#####7    | 450/781 [00:23<00:17, 19.03it/s]
 58%|#####7    | 452/781 [00:24<00:17, 18.92it/s]
 58%|#####8    | 454/781 [00:24<00:17, 19.00it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.86it/s]
 59%|#####8    | 458/781 [00:24<00:17, 18.94it/s]
 59%|#####8    | 460/781 [00:24<00:16, 18.95it/s]
 59%|#####9    | 462/781 [00:24<00:16, 18.98it/s]
 59%|#####9    | 464/781 [00:24<00:16, 18.97it/s]
 60%|#####9    | 466/781 [00:24<00:16, 18.69it/s]
 60%|#####9    | 468/781 [00:24<00:16, 18.54it/s]
 60%|######    | 470/781 [00:25<00:16, 18.37it/s]
 60%|######    | 472/781 [00:25<00:16, 18.48it/s]
 61%|######    | 474/781 [00:25<00:16, 18.43it/s]
 61%|######    | 476/781 [00:25<00:16, 18.63it/s]
 61%|######1   | 478/781 [00:25<00:16, 18.64it/s]
 61%|######1   | 480/781 [00:25<00:16, 18.77it/s]
 62%|######1   | 482/781 [00:25<00:16, 18.56it/s]
 62%|######1   | 484/781 [00:25<00:16, 18.31it/s]
 62%|######2   | 486/781 [00:25<00:16, 18.32it/s]
 62%|######2   | 488/781 [00:26<00:15, 18.52it/s]
 63%|######2   | 490/781 [00:26<00:15, 18.36it/s]
 63%|######2   | 492/781 [00:26<00:15, 18.44it/s]
 63%|######3   | 494/781 [00:26<00:15, 18.45it/s]
 64%|######3   | 496/781 [00:26<00:15, 18.53it/s]
 64%|######3   | 498/781 [00:26<00:15, 18.48it/s]
 64%|######4   | 500/781 [00:26<00:15, 18.33it/s]
 64%|######4   | 502/781 [00:26<00:15, 18.53it/s]
 65%|######4   | 504/781 [00:26<00:14, 18.70it/s]
 65%|######4   | 506/781 [00:26<00:14, 18.60it/s]
 65%|######5   | 508/781 [00:27<00:14, 18.67it/s]
 65%|######5   | 510/781 [00:27<00:14, 18.68it/s]
 66%|######5   | 512/781 [00:27<00:14, 18.63it/s]
 66%|######5   | 514/781 [00:27<00:14, 18.38it/s]
 66%|######6   | 516/781 [00:27<00:14, 18.46it/s]
 66%|######6   | 518/781 [00:27<00:14, 18.49it/s]
 67%|######6   | 520/781 [00:27<00:14, 18.42it/s]
 67%|######6   | 522/781 [00:27<00:13, 18.63it/s]
 67%|######7   | 524/781 [00:27<00:13, 18.57it/s]
 67%|######7   | 526/781 [00:28<00:13, 18.80it/s]
 68%|######7   | 528/781 [00:28<00:13, 18.97it/s]
 68%|######7   | 530/781 [00:28<00:13, 18.86it/s]
 68%|######8   | 532/781 [00:28<00:13, 19.04it/s]
 68%|######8   | 534/781 [00:28<00:12, 19.22it/s]
 69%|######8   | 536/781 [00:28<00:12, 19.28it/s]
 69%|######8   | 538/781 [00:28<00:12, 19.34it/s]
 69%|######9   | 540/781 [00:28<00:12, 19.33it/s]
 69%|######9   | 542/781 [00:28<00:12, 19.23it/s]
 70%|######9   | 544/781 [00:28<00:12, 19.26it/s]
 70%|######9   | 546/781 [00:29<00:12, 19.26it/s]
 70%|#######   | 548/781 [00:29<00:12, 19.29it/s]
 70%|#######   | 550/781 [00:29<00:12, 19.13it/s]
 71%|#######   | 552/781 [00:29<00:12, 19.01it/s]
 71%|#######   | 554/781 [00:29<00:11, 18.97it/s]
 71%|#######1  | 556/781 [00:29<00:12, 18.35it/s]
 71%|#######1  | 558/781 [00:29<00:12, 18.35it/s]
 72%|#######1  | 560/781 [00:29<00:12, 18.12it/s]
 72%|#######1  | 562/781 [00:29<00:12, 18.22it/s]
 72%|#######2  | 564/781 [00:30<00:11, 18.41it/s]
 72%|#######2  | 566/781 [00:30<00:11, 18.37it/s]
 73%|#######2  | 568/781 [00:30<00:11, 18.16it/s]
 73%|#######2  | 570/781 [00:30<00:11, 18.04it/s]
 73%|#######3  | 572/781 [00:30<00:11, 18.05it/s]
 73%|#######3  | 574/781 [00:30<00:11, 18.11it/s]
 74%|#######3  | 576/781 [00:30<00:11, 18.32it/s]
 74%|#######4  | 578/781 [00:30<00:10, 18.61it/s]
 74%|#######4  | 580/781 [00:30<00:10, 18.83it/s]
 75%|#######4  | 582/781 [00:31<00:10, 18.99it/s]
 75%|#######4  | 584/781 [00:31<00:10, 18.96it/s]
 75%|#######5  | 586/781 [00:31<00:10, 18.94it/s]
 75%|#######5  | 588/781 [00:31<00:10, 18.76it/s]
 76%|#######5  | 590/781 [00:31<00:10, 18.98it/s]
 76%|#######5  | 592/781 [00:31<00:09, 19.04it/s]
 76%|#######6  | 594/781 [00:31<00:09, 19.08it/s]
 76%|#######6  | 596/781 [00:31<00:09, 19.00it/s]
 77%|#######6  | 598/781 [00:31<00:09, 19.14it/s]
 77%|#######6  | 600/781 [00:31<00:09, 19.23it/s]
 77%|#######7  | 602/781 [00:32<00:09, 19.33it/s]
 77%|#######7  | 604/781 [00:32<00:09, 19.42it/s]
 78%|#######7  | 606/781 [00:32<00:08, 19.46it/s]
 78%|#######7  | 608/781 [00:32<00:08, 19.38it/s]
 78%|#######8  | 610/781 [00:32<00:08, 19.27it/s]
 78%|#######8  | 612/781 [00:32<00:08, 18.86it/s]
 79%|#######8  | 614/781 [00:32<00:08, 19.01it/s]
 79%|#######8  | 616/781 [00:32<00:08, 18.94it/s]
 79%|#######9  | 618/781 [00:32<00:08, 18.88it/s]
 79%|#######9  | 620/781 [00:33<00:08, 18.81it/s]
 80%|#######9  | 622/781 [00:33<00:08, 18.62it/s]
 80%|#######9  | 624/781 [00:33<00:08, 18.44it/s]
 80%|########  | 626/781 [00:33<00:08, 18.70it/s]
 80%|########  | 628/781 [00:33<00:08, 18.74it/s]
 81%|########  | 630/781 [00:33<00:07, 18.90it/s]
 81%|########  | 632/781 [00:33<00:07, 18.83it/s]
 81%|########1 | 634/781 [00:33<00:07, 18.82it/s]
 81%|########1 | 636/781 [00:33<00:07, 18.76it/s]
 82%|########1 | 638/781 [00:34<00:07, 18.62it/s]
 82%|########1 | 640/781 [00:34<00:07, 18.62it/s]
 82%|########2 | 642/781 [00:34<00:07, 18.30it/s]
 82%|########2 | 644/781 [00:34<00:07, 18.22it/s]
 83%|########2 | 646/781 [00:34<00:07, 18.21it/s]
 83%|########2 | 648/781 [00:34<00:07, 18.19it/s]
 83%|########3 | 650/781 [00:34<00:07, 18.18it/s]
 83%|########3 | 652/781 [00:34<00:07, 18.40it/s]
 84%|########3 | 654/781 [00:34<00:06, 18.53it/s]
 84%|########3 | 656/781 [00:34<00:06, 18.71it/s]
 84%|########4 | 658/781 [00:35<00:06, 18.85it/s]
 85%|########4 | 660/781 [00:35<00:06, 18.92it/s]
 85%|########4 | 662/781 [00:35<00:06, 19.00it/s]
 85%|########5 | 664/781 [00:35<00:06, 19.01it/s]
 85%|########5 | 666/781 [00:35<00:06, 18.98it/s]
 86%|########5 | 668/781 [00:35<00:06, 18.80it/s]
 86%|########5 | 670/781 [00:35<00:05, 18.75it/s]
 86%|########6 | 672/781 [00:35<00:05, 18.83it/s]
 86%|########6 | 674/781 [00:35<00:05, 18.91it/s]
 87%|########6 | 676/781 [00:36<00:05, 18.84it/s]
 87%|########6 | 678/781 [00:36<00:05, 18.89it/s]
 87%|########7 | 680/781 [00:36<00:05, 18.81it/s]
 87%|########7 | 682/781 [00:36<00:05, 18.79it/s]
 88%|########7 | 684/781 [00:36<00:05, 18.72it/s]
 88%|########7 | 686/781 [00:36<00:05, 18.62it/s]
 88%|########8 | 688/781 [00:36<00:05, 18.59it/s]
 88%|########8 | 690/781 [00:36<00:04, 18.61it/s]
 89%|########8 | 692/781 [00:36<00:04, 18.77it/s]
 89%|########8 | 694/781 [00:37<00:04, 18.85it/s]
 89%|########9 | 696/781 [00:37<00:04, 18.95it/s]
 89%|########9 | 698/781 [00:37<00:04, 18.96it/s]
 90%|########9 | 700/781 [00:37<00:04, 19.02it/s]
 90%|########9 | 702/781 [00:37<00:04, 19.05it/s]
 90%|######### | 704/781 [00:37<00:04, 19.12it/s]
 90%|######### | 706/781 [00:37<00:03, 18.98it/s]
 91%|######### | 708/781 [00:37<00:03, 19.00it/s]
 91%|######### | 710/781 [00:37<00:03, 18.99it/s]
 91%|#########1| 712/781 [00:37<00:03, 19.04it/s]
 91%|#########1| 714/781 [00:38<00:03, 19.14it/s]
 92%|#########1| 716/781 [00:38<00:03, 19.16it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.14it/s]
 92%|#########2| 720/781 [00:38<00:03, 19.04it/s]
 92%|#########2| 722/781 [00:38<00:03, 18.97it/s]
 93%|#########2| 724/781 [00:38<00:03, 18.80it/s]
 93%|#########2| 726/781 [00:38<00:02, 18.75it/s]
 93%|#########3| 728/781 [00:38<00:02, 18.70it/s]
 93%|#########3| 730/781 [00:38<00:02, 18.70it/s]
 94%|#########3| 732/781 [00:39<00:02, 18.60it/s]
 94%|#########3| 734/781 [00:39<00:02, 18.59it/s]
 94%|#########4| 736/781 [00:39<00:02, 18.66it/s]
 94%|#########4| 738/781 [00:39<00:02, 18.39it/s]
 95%|#########4| 740/781 [00:39<00:02, 18.59it/s]
 95%|#########5| 742/781 [00:39<00:02, 18.69it/s]
 95%|#########5| 744/781 [00:39<00:01, 18.78it/s]
 96%|#########5| 746/781 [00:39<00:01, 18.78it/s]
 96%|#########5| 748/781 [00:39<00:01, 18.87it/s]
 96%|#########6| 750/781 [00:39<00:01, 18.68it/s]
 96%|#########6| 752/781 [00:40<00:01, 18.49it/s]
 97%|#########6| 754/781 [00:40<00:01, 18.42it/s]
 97%|#########6| 756/781 [00:40<00:01, 18.32it/s]
 97%|#########7| 758/781 [00:40<00:01, 18.24it/s]
 97%|#########7| 760/781 [00:40<00:01, 18.20it/s]
 98%|#########7| 762/781 [00:40<00:01, 18.36it/s]
 98%|#########7| 764/781 [00:40<00:00, 18.52it/s]
 98%|#########8| 766/781 [00:40<00:00, 18.51it/s]
 98%|#########8| 768/781 [00:40<00:00, 18.65it/s]
 99%|#########8| 770/781 [00:41<00:00, 18.55it/s]
 99%|#########8| 772/781 [00:41<00:00, 18.69it/s]
 99%|#########9| 774/781 [00:41<00:00, 18.79it/s]
 99%|#########9| 776/781 [00:41<00:00, 18.87it/s]
100%|#########9| 778/781 [00:41<00:00, 18.87it/s]
100%|#########9| 780/781 [00:41<00:00, 18.88it/s]
100%|##########| 781/781 [00:41<00:00, 18.75it/s]








Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.15s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:48,  1.16s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:50,  1.20s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.27s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.25s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:46,  1.20s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:41,  1.20s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.25s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:40,  1.26s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.26s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:39,  1.27s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:36,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:35,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.29s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:35,  1.30s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.29s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:29,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.25s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.28s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.33s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.14s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.27s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:04<00:00,  4.93s/it]
100%|##########| 1/1 [00:04<00:00,  4.94s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:12, 12.30it/s]
  3%|2         | 4/157 [00:00<00:11, 12.81it/s]
  4%|3         | 6/157 [00:00<00:11, 13.54it/s]
  5%|5         | 8/157 [00:00<00:10, 14.25it/s]
  6%|6         | 10/157 [00:00<00:09, 14.73it/s]
  8%|7         | 12/157 [00:00<00:09, 14.99it/s]
  9%|8         | 14/157 [00:00<00:09, 15.18it/s]
 10%|#         | 16/157 [00:01<00:09, 15.18it/s]
 11%|#1        | 18/157 [00:01<00:09, 15.26it/s]
 13%|#2        | 20/157 [00:01<00:08, 15.45it/s]
 14%|#4        | 22/157 [00:01<00:08, 15.48it/s]
 15%|#5        | 24/157 [00:01<00:08, 15.52it/s]
 17%|#6        | 26/157 [00:01<00:08, 15.53it/s]
 18%|#7        | 28/157 [00:01<00:08, 15.67it/s]
 19%|#9        | 30/157 [00:01<00:08, 15.70it/s]
 20%|##        | 32/157 [00:02<00:08, 15.62it/s]
 22%|##1       | 34/157 [00:02<00:07, 15.57it/s]
 23%|##2       | 36/157 [00:02<00:07, 15.55it/s]
 24%|##4       | 38/157 [00:02<00:07, 15.64it/s]
 25%|##5       | 40/157 [00:02<00:07, 15.64it/s]
 27%|##6       | 42/157 [00:02<00:07, 15.69it/s]
 28%|##8       | 44/157 [00:02<00:07, 15.70it/s]
 29%|##9       | 46/157 [00:03<00:07, 15.72it/s]
 31%|###       | 48/157 [00:03<00:06, 15.79it/s]
 32%|###1      | 50/157 [00:03<00:06, 15.83it/s]
 33%|###3      | 52/157 [00:03<00:06, 15.76it/s]
 34%|###4      | 54/157 [00:03<00:06, 15.76it/s]
 36%|###5      | 56/157 [00:03<00:06, 15.64it/s]
 37%|###6      | 58/157 [00:03<00:06, 15.78it/s]
 38%|###8      | 60/157 [00:03<00:06, 15.62it/s]
 39%|###9      | 62/157 [00:04<00:06, 15.55it/s]
 41%|####      | 64/157 [00:04<00:05, 15.59it/s]
 42%|####2     | 66/157 [00:04<00:05, 15.68it/s]
 43%|####3     | 68/157 [00:04<00:05, 15.67it/s]
 45%|####4     | 70/157 [00:04<00:05, 15.74it/s]
 46%|####5     | 72/157 [00:04<00:05, 15.78it/s]
 47%|####7     | 74/157 [00:04<00:05, 15.84it/s]
 48%|####8     | 76/157 [00:04<00:05, 15.84it/s]
 50%|####9     | 78/157 [00:05<00:05, 15.63it/s]
 51%|#####     | 80/157 [00:05<00:04, 15.63it/s]
 52%|#####2    | 82/157 [00:05<00:04, 15.64it/s]
 54%|#####3    | 84/157 [00:05<00:04, 15.66it/s]
 55%|#####4    | 86/157 [00:05<00:04, 15.63it/s]
 56%|#####6    | 88/157 [00:05<00:04, 15.63it/s]
 57%|#####7    | 90/157 [00:05<00:04, 15.71it/s]
 59%|#####8    | 92/157 [00:05<00:04, 15.80it/s]
 60%|#####9    | 94/157 [00:06<00:03, 15.84it/s]
 61%|######1   | 96/157 [00:06<00:03, 15.78it/s]
 62%|######2   | 98/157 [00:06<00:03, 15.72it/s]
 64%|######3   | 100/157 [00:06<00:03, 15.65it/s]
 65%|######4   | 102/157 [00:06<00:03, 15.71it/s]
 66%|######6   | 104/157 [00:06<00:03, 15.64it/s]
 68%|######7   | 106/157 [00:06<00:03, 15.55it/s]
 69%|######8   | 108/157 [00:06<00:03, 15.64it/s]
 70%|#######   | 110/157 [00:07<00:02, 15.77it/s]
 71%|#######1  | 112/157 [00:07<00:02, 15.82it/s]
 73%|#######2  | 114/157 [00:07<00:02, 15.81it/s]
 74%|#######3  | 116/157 [00:07<00:02, 15.83it/s]
 75%|#######5  | 118/157 [00:07<00:02, 15.76it/s]
 76%|#######6  | 120/157 [00:07<00:02, 15.71it/s]
 78%|#######7  | 122/157 [00:07<00:02, 15.76it/s]
 79%|#######8  | 124/157 [00:07<00:02, 15.82it/s]
 80%|########  | 126/157 [00:08<00:01, 15.80it/s]
 82%|########1 | 128/157 [00:08<00:01, 15.75it/s]
 83%|########2 | 130/157 [00:08<00:01, 15.72it/s]
 84%|########4 | 132/157 [00:08<00:01, 15.56it/s]
 85%|########5 | 134/157 [00:08<00:01, 15.71it/s]
 87%|########6 | 136/157 [00:08<00:01, 15.79it/s]
 88%|########7 | 138/157 [00:08<00:01, 15.79it/s]
 89%|########9 | 140/157 [00:08<00:01, 15.72it/s]
 90%|######### | 142/157 [00:09<00:00, 15.57it/s]
 92%|#########1| 144/157 [00:09<00:00, 15.38it/s]
 93%|#########2| 146/157 [00:09<00:00, 15.29it/s]
 94%|#########4| 148/157 [00:09<00:00, 15.24it/s]
 96%|#########5| 150/157 [00:09<00:00, 15.11it/s]
 97%|#########6| 152/157 [00:09<00:00, 15.36it/s]
 98%|#########8| 154/157 [00:09<00:00, 15.55it/s]
 99%|#########9| 156/157 [00:10<00:00, 15.64it/s]
100%|##########| 157/157 [00:10<00:00, 15.60it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:49, 15.70it/s]
  1%|          | 4/781 [00:00<00:47, 16.21it/s]
  1%|          | 6/781 [00:00<00:48, 15.95it/s]
  1%|1         | 8/781 [00:00<00:48, 16.04it/s]
  1%|1         | 10/781 [00:00<00:47, 16.25it/s]
  2%|1         | 12/781 [00:00<00:47, 16.34it/s]
  2%|1         | 14/781 [00:00<00:45, 16.82it/s]
  2%|2         | 16/781 [00:00<00:45, 16.94it/s]
  2%|2         | 18/781 [00:01<00:45, 16.91it/s]
  3%|2         | 20/781 [00:01<00:44, 17.17it/s]
  3%|2         | 22/781 [00:01<00:44, 17.25it/s]
  3%|3         | 24/781 [00:01<00:45, 16.74it/s]
  3%|3         | 26/781 [00:01<00:44, 17.14it/s]
  4%|3         | 28/781 [00:01<00:43, 17.38it/s]
  4%|3         | 30/781 [00:01<00:42, 17.48it/s]
  4%|4         | 32/781 [00:01<00:43, 17.26it/s]
  4%|4         | 34/781 [00:01<00:41, 17.85it/s]
  5%|4         | 36/781 [00:02<00:41, 17.77it/s]
  5%|4         | 38/781 [00:02<00:42, 17.67it/s]
  5%|5         | 40/781 [00:02<00:42, 17.55it/s]
  5%|5         | 42/781 [00:02<00:42, 17.53it/s]
  6%|5         | 44/781 [00:02<00:42, 17.50it/s]
  6%|5         | 46/781 [00:02<00:42, 17.50it/s]
  6%|6         | 48/781 [00:02<00:42, 17.43it/s]
  6%|6         | 50/781 [00:02<00:41, 17.47it/s]
  7%|6         | 52/781 [00:03<00:41, 17.54it/s]
  7%|6         | 54/781 [00:03<00:41, 17.69it/s]
  7%|7         | 56/781 [00:03<00:40, 17.76it/s]
  7%|7         | 58/781 [00:03<00:40, 17.80it/s]
  8%|7         | 60/781 [00:03<00:40, 17.68it/s]
  8%|7         | 62/781 [00:03<00:40, 17.83it/s]
  8%|8         | 64/781 [00:03<00:39, 17.94it/s]
  8%|8         | 66/781 [00:03<00:39, 18.01it/s]
  9%|8         | 68/781 [00:03<00:39, 17.93it/s]
  9%|8         | 70/781 [00:04<00:39, 18.04it/s]
  9%|9         | 72/781 [00:04<00:39, 18.05it/s]
  9%|9         | 74/781 [00:04<00:39, 18.09it/s]
 10%|9         | 76/781 [00:04<00:38, 18.14it/s]
 10%|9         | 78/781 [00:04<00:38, 18.15it/s]
 10%|#         | 80/781 [00:04<00:38, 18.06it/s]
 10%|#         | 82/781 [00:04<00:38, 18.05it/s]
 11%|#         | 84/781 [00:04<00:38, 17.91it/s]
 11%|#1        | 86/781 [00:04<00:38, 17.98it/s]
 11%|#1        | 88/781 [00:05<00:38, 18.04it/s]
 12%|#1        | 90/781 [00:05<00:38, 18.06it/s]
 12%|#1        | 92/781 [00:05<00:38, 18.11it/s]
 12%|#2        | 94/781 [00:05<00:37, 18.11it/s]
 12%|#2        | 96/781 [00:05<00:37, 18.16it/s]
 13%|#2        | 98/781 [00:05<00:37, 18.20it/s]
 13%|#2        | 100/781 [00:05<00:37, 18.19it/s]
 13%|#3        | 102/781 [00:05<00:37, 18.15it/s]
 13%|#3        | 104/781 [00:05<00:37, 18.16it/s]
 14%|#3        | 106/781 [00:06<00:37, 18.07it/s]
 14%|#3        | 108/781 [00:06<00:37, 18.01it/s]
 14%|#4        | 110/781 [00:06<00:37, 17.95it/s]
 14%|#4        | 112/781 [00:06<00:37, 17.80it/s]
 15%|#4        | 114/781 [00:06<00:37, 17.90it/s]
 15%|#4        | 116/781 [00:06<00:37, 17.95it/s]
 15%|#5        | 118/781 [00:06<00:36, 17.99it/s]
 15%|#5        | 120/781 [00:06<00:36, 17.91it/s]
 16%|#5        | 122/781 [00:06<00:36, 18.01it/s]
 16%|#5        | 124/781 [00:07<00:36, 18.08it/s]
 16%|#6        | 126/781 [00:07<00:36, 18.08it/s]
 16%|#6        | 128/781 [00:07<00:36, 18.01it/s]
 17%|#6        | 130/781 [00:07<00:36, 17.79it/s]
 17%|#6        | 132/781 [00:07<00:36, 17.62it/s]
 17%|#7        | 134/781 [00:07<00:36, 17.76it/s]
 17%|#7        | 136/781 [00:07<00:36, 17.87it/s]
 18%|#7        | 138/781 [00:07<00:35, 17.87it/s]
 18%|#7        | 140/781 [00:07<00:36, 17.79it/s]
 18%|#8        | 142/781 [00:08<00:36, 17.57it/s]
 18%|#8        | 144/781 [00:08<00:36, 17.51it/s]
 19%|#8        | 146/781 [00:08<00:35, 17.65it/s]
 19%|#8        | 148/781 [00:08<00:35, 17.68it/s]
 19%|#9        | 150/781 [00:08<00:35, 17.81it/s]
 19%|#9        | 152/781 [00:08<00:35, 17.89it/s]
 20%|#9        | 154/781 [00:08<00:34, 18.00it/s]
 20%|#9        | 156/781 [00:08<00:34, 18.04it/s]
 20%|##        | 158/781 [00:08<00:34, 18.08it/s]
 20%|##        | 160/781 [00:09<00:34, 17.93it/s]
 21%|##        | 162/781 [00:09<00:34, 17.97it/s]
 21%|##        | 164/781 [00:09<00:34, 17.95it/s]
 21%|##1       | 166/781 [00:09<00:34, 18.02it/s]
 22%|##1       | 168/781 [00:09<00:34, 17.75it/s]
 22%|##1       | 170/781 [00:09<00:34, 17.86it/s]
 22%|##2       | 172/781 [00:09<00:34, 17.85it/s]
 22%|##2       | 174/781 [00:09<00:33, 17.89it/s]
 23%|##2       | 176/781 [00:09<00:33, 17.84it/s]
 23%|##2       | 178/781 [00:10<00:33, 17.82it/s]
 23%|##3       | 180/781 [00:10<00:33, 17.68it/s]
 23%|##3       | 182/781 [00:10<00:33, 17.79it/s]
 24%|##3       | 184/781 [00:10<00:33, 17.64it/s]
 24%|##3       | 186/781 [00:10<00:33, 17.55it/s]
 24%|##4       | 188/781 [00:10<00:33, 17.65it/s]
 24%|##4       | 190/781 [00:10<00:33, 17.72it/s]
 25%|##4       | 192/781 [00:10<00:33, 17.78it/s]
 25%|##4       | 194/781 [00:10<00:32, 17.80it/s]
 25%|##5       | 196/781 [00:11<00:32, 17.87it/s]
 25%|##5       | 198/781 [00:11<00:32, 17.92it/s]
 26%|##5       | 200/781 [00:11<00:32, 17.79it/s]
 26%|##5       | 202/781 [00:11<00:32, 17.73it/s]
 26%|##6       | 204/781 [00:11<00:32, 17.80it/s]
 26%|##6       | 206/781 [00:11<00:32, 17.93it/s]
 27%|##6       | 208/781 [00:11<00:31, 17.93it/s]
 27%|##6       | 210/781 [00:11<00:31, 17.96it/s]
 27%|##7       | 212/781 [00:11<00:31, 18.06it/s]
 27%|##7       | 214/781 [00:12<00:31, 18.11it/s]
 28%|##7       | 216/781 [00:12<00:31, 18.12it/s]
 28%|##7       | 218/781 [00:12<00:30, 18.17it/s]
 28%|##8       | 220/781 [00:12<00:31, 17.85it/s]
 28%|##8       | 222/781 [00:12<00:31, 17.97it/s]
 29%|##8       | 224/781 [00:12<00:31, 17.93it/s]
 29%|##8       | 226/781 [00:12<00:30, 17.98it/s]
 29%|##9       | 228/781 [00:12<00:30, 18.06it/s]
 29%|##9       | 230/781 [00:12<00:30, 18.12it/s]
 30%|##9       | 232/781 [00:13<00:30, 18.16it/s]
 30%|##9       | 234/781 [00:13<00:30, 18.02it/s]
 30%|###       | 236/781 [00:13<00:30, 18.04it/s]
 30%|###       | 238/781 [00:13<00:30, 18.02it/s]
 31%|###       | 240/781 [00:13<00:30, 17.94it/s]
 31%|###       | 242/781 [00:13<00:30, 17.93it/s]
 31%|###1      | 244/781 [00:13<00:29, 18.00it/s]
 31%|###1      | 246/781 [00:13<00:29, 18.04it/s]
 32%|###1      | 248/781 [00:13<00:29, 17.99it/s]
 32%|###2      | 250/781 [00:14<00:29, 17.71it/s]
 32%|###2      | 252/781 [00:14<00:29, 17.94it/s]
 33%|###2      | 254/781 [00:14<00:29, 18.04it/s]
 33%|###2      | 256/781 [00:14<00:29, 17.97it/s]
 33%|###3      | 258/781 [00:14<00:29, 17.98it/s]
 33%|###3      | 260/781 [00:14<00:28, 18.39it/s]
 34%|###3      | 262/781 [00:14<00:28, 18.47it/s]
 34%|###3      | 264/781 [00:14<00:27, 18.81it/s]
 34%|###4      | 266/781 [00:14<00:27, 18.45it/s]
 34%|###4      | 268/781 [00:15<00:27, 18.51it/s]
 35%|###4      | 270/781 [00:15<00:27, 18.30it/s]
 35%|###4      | 272/781 [00:15<00:27, 18.28it/s]
 35%|###5      | 274/781 [00:15<00:27, 18.25it/s]
 35%|###5      | 276/781 [00:15<00:27, 18.22it/s]
 36%|###5      | 278/781 [00:15<00:27, 18.28it/s]
 36%|###5      | 280/781 [00:15<00:27, 18.15it/s]
 36%|###6      | 282/781 [00:15<00:27, 17.96it/s]
 36%|###6      | 284/781 [00:15<00:28, 17.73it/s]
 37%|###6      | 286/781 [00:16<00:27, 17.86it/s]
 37%|###6      | 288/781 [00:16<00:27, 17.99it/s]
 37%|###7      | 290/781 [00:16<00:27, 17.99it/s]
 37%|###7      | 292/781 [00:16<00:27, 18.04it/s]
 38%|###7      | 294/781 [00:16<00:27, 17.94it/s]
 38%|###7      | 296/781 [00:16<00:26, 17.98it/s]
 38%|###8      | 298/781 [00:16<00:26, 18.06it/s]
 38%|###8      | 300/781 [00:16<00:26, 18.09it/s]
 39%|###8      | 302/781 [00:16<00:26, 18.06it/s]
 39%|###8      | 304/781 [00:17<00:26, 17.89it/s]
 39%|###9      | 306/781 [00:17<00:26, 17.92it/s]
 39%|###9      | 308/781 [00:17<00:26, 17.90it/s]
 40%|###9      | 310/781 [00:17<00:26, 17.91it/s]
 40%|###9      | 312/781 [00:17<00:26, 17.90it/s]
 40%|####      | 314/781 [00:17<00:26, 17.82it/s]
 40%|####      | 316/781 [00:17<00:26, 17.76it/s]
 41%|####      | 318/781 [00:17<00:26, 17.76it/s]
 41%|####      | 320/781 [00:17<00:25, 17.76it/s]
 41%|####1     | 322/781 [00:18<00:25, 17.67it/s]
 41%|####1     | 324/781 [00:18<00:25, 17.70it/s]
 42%|####1     | 326/781 [00:18<00:25, 17.67it/s]
 42%|####1     | 328/781 [00:18<00:25, 17.62it/s]
 42%|####2     | 330/781 [00:18<00:25, 17.51it/s]
 43%|####2     | 332/781 [00:18<00:25, 17.47it/s]
 43%|####2     | 334/781 [00:18<00:25, 17.70it/s]
 43%|####3     | 336/781 [00:18<00:25, 17.80it/s]
 43%|####3     | 338/781 [00:18<00:25, 17.55it/s]
 44%|####3     | 340/781 [00:19<00:24, 17.75it/s]
 44%|####3     | 342/781 [00:19<00:24, 17.90it/s]
 44%|####4     | 344/781 [00:19<00:24, 17.97it/s]
 44%|####4     | 346/781 [00:19<00:24, 18.03it/s]
 45%|####4     | 348/781 [00:19<00:23, 18.04it/s]
 45%|####4     | 350/781 [00:19<00:24, 17.95it/s]
 45%|####5     | 352/781 [00:19<00:24, 17.86it/s]
 45%|####5     | 354/781 [00:19<00:23, 18.00it/s]
 46%|####5     | 356/781 [00:19<00:23, 17.96it/s]
 46%|####5     | 358/781 [00:20<00:23, 17.87it/s]
 46%|####6     | 360/781 [00:20<00:23, 17.95it/s]
 46%|####6     | 362/781 [00:20<00:23, 17.93it/s]
 47%|####6     | 364/781 [00:20<00:23, 17.80it/s]
 47%|####6     | 366/781 [00:20<00:23, 17.73it/s]
 47%|####7     | 368/781 [00:20<00:23, 17.82it/s]
 47%|####7     | 370/781 [00:20<00:23, 17.59it/s]
 48%|####7     | 372/781 [00:20<00:23, 17.71it/s]
 48%|####7     | 374/781 [00:20<00:22, 17.78it/s]
 48%|####8     | 376/781 [00:21<00:22, 17.76it/s]
 48%|####8     | 378/781 [00:21<00:22, 17.75it/s]
 49%|####8     | 380/781 [00:21<00:22, 17.79it/s]
 49%|####8     | 382/781 [00:21<00:22, 17.68it/s]
 49%|####9     | 384/781 [00:21<00:22, 17.79it/s]
 49%|####9     | 386/781 [00:21<00:22, 17.74it/s]
 50%|####9     | 388/781 [00:21<00:22, 17.67it/s]
 50%|####9     | 390/781 [00:21<00:22, 17.37it/s]
 50%|#####     | 392/781 [00:22<00:22, 17.59it/s]
 50%|#####     | 394/781 [00:22<00:21, 17.75it/s]
 51%|#####     | 396/781 [00:22<00:21, 17.51it/s]
 51%|#####     | 398/781 [00:22<00:21, 17.73it/s]
 51%|#####1    | 400/781 [00:22<00:21, 17.82it/s]
 51%|#####1    | 402/781 [00:22<00:21, 17.85it/s]
 52%|#####1    | 404/781 [00:22<00:21, 17.90it/s]
 52%|#####1    | 406/781 [00:22<00:20, 18.02it/s]
 52%|#####2    | 408/781 [00:22<00:20, 18.11it/s]
 52%|#####2    | 410/781 [00:23<00:20, 18.20it/s]
 53%|#####2    | 412/781 [00:23<00:20, 18.22it/s]
 53%|#####3    | 414/781 [00:23<00:20, 18.24it/s]
 53%|#####3    | 416/781 [00:23<00:19, 18.26it/s]
 54%|#####3    | 418/781 [00:23<00:19, 18.22it/s]
 54%|#####3    | 420/781 [00:23<00:19, 18.17it/s]
 54%|#####4    | 422/781 [00:23<00:19, 18.11it/s]
 54%|#####4    | 424/781 [00:23<00:19, 18.04it/s]
 55%|#####4    | 426/781 [00:23<00:19, 18.00it/s]
 55%|#####4    | 428/781 [00:23<00:19, 18.02it/s]
 55%|#####5    | 430/781 [00:24<00:19, 18.09it/s]
 55%|#####5    | 432/781 [00:24<00:19, 18.10it/s]
 56%|#####5    | 434/781 [00:24<00:19, 18.11it/s]
 56%|#####5    | 436/781 [00:24<00:19, 17.91it/s]
 56%|#####6    | 438/781 [00:24<00:19, 17.90it/s]
 56%|#####6    | 440/781 [00:24<00:19, 17.84it/s]
 57%|#####6    | 442/781 [00:24<00:18, 17.91it/s]
 57%|#####6    | 444/781 [00:24<00:18, 18.05it/s]
 57%|#####7    | 446/781 [00:24<00:18, 18.15it/s]
 57%|#####7    | 448/781 [00:25<00:18, 18.20it/s]
 58%|#####7    | 450/781 [00:25<00:18, 18.15it/s]
 58%|#####7    | 452/781 [00:25<00:18, 18.17it/s]
 58%|#####8    | 454/781 [00:25<00:17, 18.17it/s]
 58%|#####8    | 456/781 [00:25<00:17, 18.20it/s]
 59%|#####8    | 458/781 [00:25<00:17, 18.22it/s]
 59%|#####8    | 460/781 [00:25<00:17, 18.20it/s]
 59%|#####9    | 462/781 [00:25<00:17, 18.15it/s]
 59%|#####9    | 464/781 [00:25<00:17, 18.17it/s]
 60%|#####9    | 466/781 [00:26<00:17, 18.08it/s]
 60%|#####9    | 468/781 [00:26<00:17, 17.91it/s]
 60%|######    | 470/781 [00:26<00:17, 17.78it/s]
 60%|######    | 472/781 [00:26<00:17, 17.77it/s]
 61%|######    | 474/781 [00:26<00:17, 17.80it/s]
 61%|######    | 476/781 [00:26<00:17, 17.82it/s]
 61%|######1   | 478/781 [00:26<00:17, 17.70it/s]
 61%|######1   | 480/781 [00:26<00:17, 17.65it/s]
 62%|######1   | 482/781 [00:27<00:16, 17.68it/s]
 62%|######1   | 484/781 [00:27<00:16, 17.78it/s]
 62%|######2   | 486/781 [00:27<00:16, 17.77it/s]
 62%|######2   | 488/781 [00:27<00:16, 17.80it/s]
 63%|######2   | 490/781 [00:27<00:16, 17.85it/s]
 63%|######2   | 492/781 [00:27<00:16, 17.94it/s]
 63%|######3   | 494/781 [00:27<00:15, 17.98it/s]
 64%|######3   | 496/781 [00:27<00:16, 17.75it/s]
 64%|######3   | 498/781 [00:27<00:15, 17.87it/s]
 64%|######4   | 500/781 [00:28<00:15, 17.64it/s]
 64%|######4   | 502/781 [00:28<00:15, 17.83it/s]
 65%|######4   | 504/781 [00:28<00:15, 17.96it/s]
 65%|######4   | 506/781 [00:28<00:15, 17.98it/s]
 65%|######5   | 508/781 [00:28<00:15, 17.96it/s]
 65%|######5   | 510/781 [00:28<00:15, 17.91it/s]
 66%|######5   | 512/781 [00:28<00:15, 17.73it/s]
 66%|######5   | 514/781 [00:28<00:15, 17.78it/s]
 66%|######6   | 516/781 [00:28<00:14, 17.89it/s]
 66%|######6   | 518/781 [00:29<00:14, 17.96it/s]
 67%|######6   | 520/781 [00:29<00:14, 18.04it/s]
 67%|######6   | 522/781 [00:29<00:14, 18.10it/s]
 67%|######7   | 524/781 [00:29<00:14, 18.15it/s]
 67%|######7   | 526/781 [00:29<00:14, 18.19it/s]
 68%|######7   | 528/781 [00:29<00:13, 18.11it/s]
 68%|######7   | 530/781 [00:29<00:13, 18.02it/s]
 68%|######8   | 532/781 [00:29<00:13, 18.02it/s]
 68%|######8   | 534/781 [00:29<00:13, 18.10it/s]
 69%|######8   | 536/781 [00:30<00:13, 18.14it/s]
 69%|######8   | 538/781 [00:30<00:13, 18.17it/s]
 69%|######9   | 540/781 [00:30<00:13, 18.03it/s]
 69%|######9   | 542/781 [00:30<00:13, 17.98it/s]
 70%|######9   | 544/781 [00:30<00:13, 17.85it/s]
 70%|######9   | 546/781 [00:30<00:13, 17.85it/s]
 70%|#######   | 548/781 [00:30<00:13, 17.77it/s]
 70%|#######   | 550/781 [00:30<00:12, 17.80it/s]
 71%|#######   | 552/781 [00:30<00:12, 17.91it/s]
 71%|#######   | 554/781 [00:31<00:12, 18.01it/s]
 71%|#######1  | 556/781 [00:31<00:12, 17.94it/s]
 71%|#######1  | 558/781 [00:31<00:12, 17.82it/s]
 72%|#######1  | 560/781 [00:31<00:12, 17.84it/s]
 72%|#######1  | 562/781 [00:31<00:12, 17.84it/s]
 72%|#######2  | 564/781 [00:31<00:12, 17.93it/s]
 72%|#######2  | 566/781 [00:31<00:11, 18.00it/s]
 73%|#######2  | 568/781 [00:31<00:11, 18.07it/s]
 73%|#######2  | 570/781 [00:31<00:11, 18.16it/s]
 73%|#######3  | 572/781 [00:32<00:11, 18.14it/s]
 73%|#######3  | 574/781 [00:32<00:11, 18.14it/s]
 74%|#######3  | 576/781 [00:32<00:11, 17.99it/s]
 74%|#######4  | 578/781 [00:32<00:11, 18.09it/s]
 74%|#######4  | 580/781 [00:32<00:11, 18.02it/s]
 75%|#######4  | 582/781 [00:32<00:10, 18.11it/s]
 75%|#######4  | 584/781 [00:32<00:10, 18.14it/s]
 75%|#######5  | 586/781 [00:32<00:10, 18.19it/s]
 75%|#######5  | 588/781 [00:32<00:10, 18.23it/s]
 76%|#######5  | 590/781 [00:33<00:10, 18.23it/s]
 76%|#######5  | 592/781 [00:33<00:10, 18.00it/s]
 76%|#######6  | 594/781 [00:33<00:10, 18.06it/s]
 76%|#######6  | 596/781 [00:33<00:10, 18.03it/s]
 77%|#######6  | 598/781 [00:33<00:10, 18.05it/s]
 77%|#######6  | 600/781 [00:33<00:10, 18.04it/s]
 77%|#######7  | 602/781 [00:33<00:10, 17.81it/s]
 77%|#######7  | 604/781 [00:33<00:09, 17.88it/s]
 78%|#######7  | 606/781 [00:33<00:09, 17.94it/s]
 78%|#######7  | 608/781 [00:34<00:09, 18.02it/s]
 78%|#######8  | 610/781 [00:34<00:09, 18.00it/s]
 78%|#######8  | 612/781 [00:34<00:09, 17.86it/s]
 79%|#######8  | 614/781 [00:34<00:09, 17.72it/s]
 79%|#######8  | 616/781 [00:34<00:09, 17.81it/s]
 79%|#######9  | 618/781 [00:34<00:09, 17.95it/s]
 79%|#######9  | 620/781 [00:34<00:08, 18.05it/s]
 80%|#######9  | 622/781 [00:34<00:08, 18.02it/s]
 80%|#######9  | 624/781 [00:34<00:08, 17.90it/s]
 80%|########  | 626/781 [00:35<00:08, 17.98it/s]
 80%|########  | 628/781 [00:35<00:08, 18.05it/s]
 81%|########  | 630/781 [00:35<00:08, 18.11it/s]
 81%|########  | 632/781 [00:35<00:08, 17.97it/s]
 81%|########1 | 634/781 [00:35<00:08, 18.07it/s]
 81%|########1 | 636/781 [00:35<00:08, 17.49it/s]
 82%|########1 | 638/781 [00:35<00:08, 17.50it/s]
 82%|########1 | 640/781 [00:35<00:07, 17.69it/s]
 82%|########2 | 642/781 [00:35<00:07, 17.79it/s]
 82%|########2 | 644/781 [00:36<00:07, 17.72it/s]
 83%|########2 | 646/781 [00:36<00:07, 17.90it/s]
 83%|########2 | 648/781 [00:36<00:07, 17.95it/s]
 83%|########3 | 650/781 [00:36<00:07, 18.01it/s]
 83%|########3 | 652/781 [00:36<00:07, 18.07it/s]
 84%|########3 | 654/781 [00:36<00:07, 18.12it/s]
 84%|########3 | 656/781 [00:36<00:06, 18.07it/s]
 84%|########4 | 658/781 [00:36<00:06, 17.83it/s]
 85%|########4 | 660/781 [00:36<00:06, 17.86it/s]
 85%|########4 | 662/781 [00:37<00:06, 17.65it/s]
 85%|########5 | 664/781 [00:37<00:06, 17.69it/s]
 85%|########5 | 666/781 [00:37<00:06, 17.84it/s]
 86%|########5 | 668/781 [00:37<00:06, 17.90it/s]
 86%|########5 | 670/781 [00:37<00:06, 17.95it/s]
 86%|########6 | 672/781 [00:37<00:06, 18.02it/s]
 86%|########6 | 674/781 [00:37<00:05, 18.12it/s]
 87%|########6 | 676/781 [00:37<00:05, 18.18it/s]
 87%|########6 | 678/781 [00:37<00:05, 18.05it/s]
 87%|########7 | 680/781 [00:38<00:05, 18.04it/s]
 87%|########7 | 682/781 [00:38<00:05, 18.11it/s]
 88%|########7 | 684/781 [00:38<00:05, 18.04it/s]
 88%|########7 | 686/781 [00:38<00:05, 18.03it/s]
 88%|########8 | 688/781 [00:38<00:05, 18.06it/s]
 88%|########8 | 690/781 [00:38<00:05, 17.89it/s]
 89%|########8 | 692/781 [00:38<00:05, 17.63it/s]
 89%|########8 | 694/781 [00:38<00:04, 17.73it/s]
 89%|########9 | 696/781 [00:38<00:04, 17.72it/s]
 89%|########9 | 698/781 [00:39<00:04, 17.65it/s]
 90%|########9 | 700/781 [00:39<00:04, 17.82it/s]
 90%|########9 | 702/781 [00:39<00:04, 17.96it/s]
 90%|######### | 704/781 [00:39<00:04, 18.06it/s]
 90%|######### | 706/781 [00:39<00:04, 18.10it/s]
 91%|######### | 708/781 [00:39<00:04, 17.90it/s]
 91%|######### | 710/781 [00:39<00:03, 17.90it/s]
 91%|#########1| 712/781 [00:39<00:03, 17.96it/s]
 91%|#########1| 714/781 [00:39<00:03, 17.94it/s]
 92%|#########1| 716/781 [00:40<00:03, 17.92it/s]
 92%|#########1| 718/781 [00:40<00:03, 18.03it/s]
 92%|#########2| 720/781 [00:40<00:03, 18.04it/s]
 92%|#########2| 722/781 [00:40<00:03, 18.02it/s]
 93%|#########2| 724/781 [00:40<00:03, 17.73it/s]
 93%|#########2| 726/781 [00:40<00:03, 17.64it/s]
 93%|#########3| 728/781 [00:40<00:03, 17.63it/s]
 93%|#########3| 730/781 [00:40<00:02, 17.68it/s]
 94%|#########3| 732/781 [00:40<00:02, 17.75it/s]
 94%|#########3| 734/781 [00:41<00:02, 18.00it/s]
 94%|#########4| 736/781 [00:41<00:02, 17.68it/s]
 94%|#########4| 738/781 [00:41<00:02, 17.75it/s]
 95%|#########4| 740/781 [00:41<00:02, 17.96it/s]
 95%|#########5| 742/781 [00:41<00:02, 17.93it/s]
 95%|#########5| 744/781 [00:41<00:02, 17.67it/s]
 96%|#########5| 746/781 [00:41<00:01, 17.68it/s]
 96%|#########5| 748/781 [00:41<00:01, 17.63it/s]
 96%|#########6| 750/781 [00:41<00:01, 17.83it/s]
 96%|#########6| 752/781 [00:42<00:01, 17.89it/s]
 97%|#########6| 754/781 [00:42<00:01, 17.92it/s]
 97%|#########6| 756/781 [00:42<00:01, 17.94it/s]
 97%|#########7| 758/781 [00:42<00:01, 17.84it/s]
 97%|#########7| 760/781 [00:42<00:01, 17.78it/s]
 98%|#########7| 762/781 [00:42<00:01, 17.93it/s]
 98%|#########7| 764/781 [00:42<00:00, 18.01it/s]
 98%|#########8| 766/781 [00:42<00:00, 18.02it/s]
 98%|#########8| 768/781 [00:42<00:00, 18.04it/s]
 99%|#########8| 770/781 [00:43<00:00, 18.09it/s]
 99%|#########8| 772/781 [00:43<00:00, 18.15it/s]
 99%|#########9| 774/781 [00:43<00:00, 18.16it/s]
 99%|#########9| 776/781 [00:43<00:00, 18.05it/s]
100%|#########9| 778/781 [00:43<00:00, 17.75it/s]
100%|#########9| 780/781 [00:43<00:00, 17.77it/s]
100%|##########| 781/781 [00:43<00:00, 17.88it/s]







Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:53,  1.25s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:46,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:46,  1.19s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.18s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.19s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:05<00:00,  5.35s/it]
100%|##########| 1/1 [00:05<00:00,  5.35s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:12, 12.84it/s]
  3%|2         | 4/157 [00:00<00:10, 14.60it/s]
  4%|3         | 6/157 [00:00<00:09, 15.57it/s]
  5%|5         | 8/157 [00:00<00:09, 16.11it/s]
  6%|6         | 10/157 [00:00<00:08, 16.43it/s]
  8%|7         | 12/157 [00:00<00:08, 16.58it/s]
  9%|8         | 14/157 [00:00<00:08, 16.65it/s]
 10%|#         | 16/157 [00:00<00:08, 16.67it/s]
 11%|#1        | 18/157 [00:01<00:08, 16.76it/s]
 13%|#2        | 20/157 [00:01<00:08, 16.77it/s]
 14%|#4        | 22/157 [00:01<00:07, 16.94it/s]
 15%|#5        | 24/157 [00:01<00:07, 16.80it/s]
 17%|#6        | 26/157 [00:01<00:07, 16.78it/s]
 18%|#7        | 28/157 [00:01<00:07, 16.61it/s]
 19%|#9        | 30/157 [00:01<00:07, 16.68it/s]
 20%|##        | 32/157 [00:01<00:07, 16.68it/s]
 22%|##1       | 34/157 [00:02<00:07, 16.75it/s]
 23%|##2       | 36/157 [00:02<00:07, 16.84it/s]
 24%|##4       | 38/157 [00:02<00:07, 16.87it/s]
 25%|##5       | 40/157 [00:02<00:06, 16.92it/s]
 27%|##6       | 42/157 [00:02<00:06, 16.90it/s]
 28%|##8       | 44/157 [00:02<00:06, 16.96it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.90it/s]
 31%|###       | 48/157 [00:02<00:06, 16.94it/s]
 32%|###1      | 50/157 [00:03<00:06, 16.99it/s]
 33%|###3      | 52/157 [00:03<00:06, 17.01it/s]
 34%|###4      | 54/157 [00:03<00:06, 17.03it/s]
 36%|###5      | 56/157 [00:03<00:05, 17.05it/s]
 37%|###6      | 58/157 [00:03<00:05, 17.07it/s]
 38%|###8      | 60/157 [00:03<00:05, 17.08it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.93it/s]
 41%|####      | 64/157 [00:03<00:05, 16.98it/s]
 42%|####2     | 66/157 [00:03<00:05, 16.85it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.83it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.78it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.45it/s]
 47%|####7     | 74/157 [00:04<00:05, 16.55it/s]
 48%|####8     | 76/157 [00:04<00:04, 16.62it/s]
 50%|####9     | 78/157 [00:04<00:04, 16.52it/s]
 51%|#####     | 80/157 [00:04<00:04, 16.50it/s]
 52%|#####2    | 82/157 [00:04<00:04, 16.68it/s]
 54%|#####3    | 84/157 [00:05<00:04, 16.49it/s]
 55%|#####4    | 86/157 [00:05<00:04, 16.32it/s]
 56%|#####6    | 88/157 [00:05<00:04, 16.45it/s]
 57%|#####7    | 90/157 [00:05<00:04, 16.49it/s]
 59%|#####8    | 92/157 [00:05<00:03, 16.46it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.34it/s]
 61%|######1   | 96/157 [00:05<00:03, 16.31it/s]
 62%|######2   | 98/157 [00:05<00:03, 16.39it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.53it/s]
 65%|######4   | 102/157 [00:06<00:03, 16.63it/s]
 66%|######6   | 104/157 [00:06<00:03, 16.77it/s]
 68%|######7   | 106/157 [00:06<00:03, 16.84it/s]
 69%|######8   | 108/157 [00:06<00:02, 16.86it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.92it/s]
 71%|#######1  | 112/157 [00:06<00:02, 16.82it/s]
 73%|#######2  | 114/157 [00:06<00:02, 16.70it/s]
 74%|#######3  | 116/157 [00:06<00:02, 16.68it/s]
 75%|#######5  | 118/157 [00:07<00:02, 16.74it/s]
 76%|#######6  | 120/157 [00:07<00:02, 16.78it/s]
 78%|#######7  | 122/157 [00:07<00:02, 16.77it/s]
 79%|#######8  | 124/157 [00:07<00:01, 16.85it/s]
 80%|########  | 126/157 [00:07<00:01, 16.90it/s]
 82%|########1 | 128/157 [00:07<00:01, 16.94it/s]
 83%|########2 | 130/157 [00:07<00:01, 16.88it/s]
 84%|########4 | 132/157 [00:07<00:01, 16.73it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.76it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.76it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.81it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.91it/s]
 90%|######### | 142/157 [00:08<00:00, 16.93it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.94it/s]
 93%|#########2| 146/157 [00:08<00:00, 16.97it/s]
 94%|#########4| 148/157 [00:08<00:00, 16.96it/s]
 96%|#########5| 150/157 [00:08<00:00, 16.96it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.81it/s]
 98%|#########8| 154/157 [00:09<00:00, 16.76it/s]
 99%|#########9| 156/157 [00:09<00:00, 16.80it/s]
100%|##########| 157/157 [00:09<00:00, 16.77it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:49, 15.85it/s]
  1%|          | 4/781 [00:00<00:44, 17.29it/s]
  1%|          | 6/781 [00:00<00:43, 17.97it/s]
  1%|1         | 8/781 [00:00<00:42, 18.36it/s]
  1%|1         | 10/781 [00:00<00:41, 18.63it/s]
  2%|1         | 12/781 [00:00<00:41, 18.75it/s]
  2%|1         | 14/781 [00:00<00:40, 18.72it/s]
  2%|2         | 16/781 [00:00<00:40, 18.78it/s]
  2%|2         | 18/781 [00:00<00:40, 18.89it/s]
  3%|2         | 20/781 [00:01<00:40, 18.98it/s]
  3%|2         | 22/781 [00:01<00:39, 19.05it/s]
  3%|3         | 24/781 [00:01<00:39, 19.07it/s]
  3%|3         | 26/781 [00:01<00:39, 18.89it/s]
  4%|3         | 28/781 [00:01<00:39, 18.92it/s]
  4%|3         | 30/781 [00:01<00:39, 19.01it/s]
  4%|4         | 32/781 [00:01<00:39, 19.02it/s]
  4%|4         | 34/781 [00:01<00:39, 19.07it/s]
  5%|4         | 36/781 [00:01<00:39, 19.10it/s]
  5%|4         | 38/781 [00:02<00:39, 19.02it/s]
  5%|5         | 40/781 [00:02<00:38, 19.07it/s]
  5%|5         | 42/781 [00:02<00:39, 18.86it/s]
  6%|5         | 44/781 [00:02<00:39, 18.81it/s]
  6%|5         | 46/781 [00:02<00:39, 18.50it/s]
  6%|6         | 48/781 [00:02<00:39, 18.38it/s]
  6%|6         | 50/781 [00:02<00:39, 18.35it/s]
  7%|6         | 52/781 [00:02<00:39, 18.26it/s]
  7%|6         | 54/781 [00:02<00:40, 18.13it/s]
  7%|7         | 56/781 [00:03<00:39, 18.34it/s]
  7%|7         | 58/781 [00:03<00:40, 18.05it/s]
  8%|7         | 60/781 [00:03<00:40, 17.90it/s]
  8%|7         | 62/781 [00:03<00:40, 17.82it/s]
  8%|8         | 64/781 [00:03<00:40, 17.85it/s]
  8%|8         | 66/781 [00:03<00:39, 18.10it/s]
  9%|8         | 68/781 [00:03<00:39, 18.22it/s]
  9%|8         | 70/781 [00:03<00:38, 18.32it/s]
  9%|9         | 72/781 [00:03<00:38, 18.33it/s]
  9%|9         | 74/781 [00:03<00:38, 18.38it/s]
 10%|9         | 76/781 [00:04<00:38, 18.45it/s]
 10%|9         | 78/781 [00:04<00:37, 18.58it/s]
 10%|#         | 80/781 [00:04<00:37, 18.66it/s]
 10%|#         | 82/781 [00:04<00:37, 18.64it/s]
 11%|#         | 84/781 [00:04<00:37, 18.75it/s]
 11%|#1        | 86/781 [00:04<00:37, 18.58it/s]
 11%|#1        | 88/781 [00:04<00:37, 18.65it/s]
 12%|#1        | 90/781 [00:04<00:36, 18.77it/s]
 12%|#1        | 92/781 [00:04<00:36, 18.65it/s]
 12%|#2        | 94/781 [00:05<00:37, 18.31it/s]
 12%|#2        | 96/781 [00:05<00:37, 18.20it/s]
 13%|#2        | 98/781 [00:05<00:37, 18.30it/s]
 13%|#2        | 100/781 [00:05<00:37, 18.23it/s]
 13%|#3        | 102/781 [00:05<00:37, 18.12it/s]
 13%|#3        | 104/781 [00:05<00:37, 17.87it/s]
 14%|#3        | 106/781 [00:05<00:37, 18.09it/s]
 14%|#3        | 108/781 [00:05<00:36, 18.35it/s]
 14%|#4        | 110/781 [00:05<00:35, 18.67it/s]
 14%|#4        | 112/781 [00:06<00:36, 18.52it/s]
 15%|#4        | 114/781 [00:06<00:35, 18.79it/s]
 15%|#4        | 116/781 [00:06<00:35, 18.83it/s]
 15%|#5        | 118/781 [00:06<00:35, 18.84it/s]
 15%|#5        | 120/781 [00:06<00:34, 18.97it/s]
 16%|#5        | 122/781 [00:06<00:34, 18.97it/s]
 16%|#5        | 124/781 [00:06<00:34, 19.09it/s]
 16%|#6        | 126/781 [00:06<00:34, 19.17it/s]
 16%|#6        | 128/781 [00:06<00:34, 19.20it/s]
 17%|#6        | 130/781 [00:06<00:33, 19.30it/s]
 17%|#6        | 132/781 [00:07<00:33, 19.40it/s]
 17%|#7        | 134/781 [00:07<00:33, 19.42it/s]
 17%|#7        | 136/781 [00:07<00:34, 18.95it/s]
 18%|#7        | 138/781 [00:07<00:33, 19.04it/s]
 18%|#7        | 140/781 [00:07<00:33, 19.06it/s]
 18%|#8        | 142/781 [00:07<00:33, 19.15it/s]
 18%|#8        | 144/781 [00:07<00:33, 19.28it/s]
 19%|#8        | 146/781 [00:07<00:32, 19.25it/s]
 19%|#8        | 148/781 [00:07<00:32, 19.25it/s]
 19%|#9        | 150/781 [00:08<00:32, 19.33it/s]
 19%|#9        | 152/781 [00:08<00:32, 19.15it/s]
 20%|#9        | 154/781 [00:08<00:32, 19.13it/s]
 20%|#9        | 156/781 [00:08<00:32, 19.11it/s]
 20%|##        | 158/781 [00:08<00:32, 19.06it/s]
 20%|##        | 160/781 [00:08<00:32, 19.07it/s]
 21%|##        | 162/781 [00:08<00:32, 19.12it/s]
 21%|##        | 164/781 [00:08<00:32, 19.18it/s]
 21%|##1       | 166/781 [00:08<00:31, 19.27it/s]
 22%|##1       | 168/781 [00:08<00:31, 19.25it/s]
 22%|##1       | 170/781 [00:09<00:31, 19.32it/s]
 22%|##2       | 172/781 [00:09<00:31, 19.35it/s]
 22%|##2       | 174/781 [00:09<00:31, 19.35it/s]
 23%|##2       | 176/781 [00:09<00:32, 18.73it/s]
 23%|##2       | 178/781 [00:09<00:31, 18.96it/s]
 23%|##3       | 180/781 [00:09<00:31, 19.11it/s]
 23%|##3       | 182/781 [00:09<00:31, 19.22it/s]
 24%|##3       | 184/781 [00:09<00:31, 19.16it/s]
 24%|##3       | 186/781 [00:09<00:31, 18.97it/s]
 24%|##4       | 188/781 [00:10<00:31, 18.93it/s]
 24%|##4       | 190/781 [00:10<00:31, 19.01it/s]
 25%|##4       | 192/781 [00:10<00:30, 19.08it/s]
 25%|##4       | 194/781 [00:10<00:30, 18.97it/s]
 25%|##5       | 196/781 [00:10<00:30, 19.08it/s]
 25%|##5       | 198/781 [00:10<00:30, 19.06it/s]
 26%|##5       | 200/781 [00:10<00:30, 18.87it/s]
 26%|##5       | 202/781 [00:10<00:30, 18.82it/s]
 26%|##6       | 204/781 [00:10<00:30, 18.85it/s]
 26%|##6       | 206/781 [00:10<00:30, 18.96it/s]
 27%|##6       | 208/781 [00:11<00:30, 19.10it/s]
 27%|##6       | 210/781 [00:11<00:29, 19.21it/s]
 27%|##7       | 212/781 [00:11<00:30, 18.92it/s]
 27%|##7       | 214/781 [00:11<00:29, 19.04it/s]
 28%|##7       | 216/781 [00:11<00:29, 18.98it/s]
 28%|##7       | 218/781 [00:11<00:29, 19.05it/s]
 28%|##8       | 220/781 [00:11<00:29, 19.13it/s]
 28%|##8       | 222/781 [00:11<00:29, 18.91it/s]
 29%|##8       | 224/781 [00:11<00:29, 19.02it/s]
 29%|##8       | 226/781 [00:12<00:29, 18.98it/s]
 29%|##9       | 228/781 [00:12<00:29, 19.07it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.97it/s]
 30%|##9       | 232/781 [00:12<00:28, 19.03it/s]
 30%|##9       | 234/781 [00:12<00:28, 19.04it/s]
 30%|###       | 236/781 [00:12<00:28, 18.95it/s]
 30%|###       | 238/781 [00:12<00:28, 19.09it/s]
 31%|###       | 240/781 [00:12<00:28, 19.19it/s]
 31%|###       | 242/781 [00:12<00:28, 19.25it/s]
 31%|###1      | 244/781 [00:12<00:27, 19.18it/s]
 31%|###1      | 246/781 [00:13<00:28, 19.11it/s]
 32%|###1      | 248/781 [00:13<00:27, 19.20it/s]
 32%|###2      | 250/781 [00:13<00:27, 19.22it/s]
 32%|###2      | 252/781 [00:13<00:27, 19.19it/s]
 33%|###2      | 254/781 [00:13<00:27, 19.17it/s]
 33%|###2      | 256/781 [00:13<00:27, 18.85it/s]
 33%|###3      | 258/781 [00:13<00:27, 18.90it/s]
 33%|###3      | 260/781 [00:13<00:27, 18.98it/s]
 34%|###3      | 262/781 [00:13<00:27, 19.05it/s]
 34%|###3      | 264/781 [00:14<00:27, 18.77it/s]
 34%|###4      | 266/781 [00:14<00:27, 18.93it/s]
 34%|###4      | 268/781 [00:14<00:27, 18.89it/s]
 35%|###4      | 270/781 [00:14<00:26, 19.02it/s]
 35%|###4      | 272/781 [00:14<00:26, 19.07it/s]
 35%|###5      | 274/781 [00:14<00:26, 19.10it/s]
 35%|###5      | 276/781 [00:14<00:26, 18.98it/s]
 36%|###5      | 278/781 [00:14<00:27, 18.62it/s]
 36%|###5      | 280/781 [00:14<00:26, 18.66it/s]
 36%|###6      | 282/781 [00:14<00:27, 18.42it/s]
 36%|###6      | 284/781 [00:15<00:27, 18.26it/s]
 37%|###6      | 286/781 [00:15<00:27, 18.06it/s]
 37%|###6      | 288/781 [00:15<00:26, 18.35it/s]
 37%|###7      | 290/781 [00:15<00:26, 18.40it/s]
 37%|###7      | 292/781 [00:15<00:26, 18.45it/s]
 38%|###7      | 294/781 [00:15<00:26, 18.39it/s]
 38%|###7      | 296/781 [00:15<00:26, 18.58it/s]
 38%|###8      | 298/781 [00:15<00:25, 18.80it/s]
 38%|###8      | 300/781 [00:15<00:25, 18.94it/s]
 39%|###8      | 302/781 [00:16<00:25, 19.03it/s]
 39%|###8      | 304/781 [00:16<00:25, 18.96it/s]
 39%|###9      | 306/781 [00:16<00:25, 18.97it/s]
 39%|###9      | 308/781 [00:16<00:24, 19.16it/s]
 40%|###9      | 310/781 [00:16<00:24, 19.18it/s]
 40%|###9      | 312/781 [00:16<00:24, 19.20it/s]
 40%|####      | 314/781 [00:16<00:24, 19.16it/s]
 40%|####      | 316/781 [00:16<00:24, 19.25it/s]
 41%|####      | 318/781 [00:16<00:24, 19.10it/s]
 41%|####      | 320/781 [00:16<00:24, 19.00it/s]
 41%|####1     | 322/781 [00:17<00:24, 19.09it/s]
 41%|####1     | 324/781 [00:17<00:23, 19.18it/s]
 42%|####1     | 326/781 [00:17<00:23, 19.14it/s]
 42%|####1     | 328/781 [00:17<00:23, 19.18it/s]
 42%|####2     | 330/781 [00:17<00:23, 19.15it/s]
 43%|####2     | 332/781 [00:17<00:23, 19.24it/s]
 43%|####2     | 334/781 [00:17<00:23, 19.31it/s]
 43%|####3     | 336/781 [00:17<00:22, 19.38it/s]
 43%|####3     | 338/781 [00:17<00:23, 19.19it/s]
 44%|####3     | 340/781 [00:18<00:22, 19.22it/s]
 44%|####3     | 342/781 [00:18<00:22, 19.29it/s]
 44%|####4     | 344/781 [00:18<00:22, 19.09it/s]
 44%|####4     | 346/781 [00:18<00:22, 19.21it/s]
 45%|####4     | 348/781 [00:18<00:22, 19.34it/s]
 45%|####4     | 350/781 [00:18<00:22, 19.09it/s]
 45%|####5     | 352/781 [00:18<00:22, 19.12it/s]
 45%|####5     | 354/781 [00:18<00:22, 19.26it/s]
 46%|####5     | 356/781 [00:18<00:22, 19.23it/s]
 46%|####5     | 358/781 [00:18<00:21, 19.28it/s]
 46%|####6     | 360/781 [00:19<00:22, 19.07it/s]
 46%|####6     | 362/781 [00:19<00:21, 19.07it/s]
 47%|####6     | 364/781 [00:19<00:21, 19.07it/s]
 47%|####6     | 366/781 [00:19<00:21, 19.14it/s]
 47%|####7     | 368/781 [00:19<00:21, 19.22it/s]
 47%|####7     | 370/781 [00:19<00:21, 19.32it/s]
 48%|####7     | 372/781 [00:19<00:21, 19.38it/s]
 48%|####7     | 374/781 [00:19<00:20, 19.46it/s]
 48%|####8     | 376/781 [00:19<00:20, 19.48it/s]
 48%|####8     | 378/781 [00:20<00:20, 19.26it/s]
 49%|####8     | 380/781 [00:20<00:20, 19.27it/s]
 49%|####8     | 382/781 [00:20<00:20, 19.23it/s]
 49%|####9     | 384/781 [00:20<00:20, 19.30it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.36it/s]
 50%|####9     | 388/781 [00:20<00:20, 19.44it/s]
 50%|####9     | 390/781 [00:20<00:20, 19.45it/s]
 50%|#####     | 392/781 [00:20<00:20, 18.95it/s]
 50%|#####     | 394/781 [00:20<00:20, 19.08it/s]
 51%|#####     | 396/781 [00:20<00:20, 19.21it/s]
 51%|#####     | 398/781 [00:21<00:19, 19.27it/s]
 51%|#####1    | 400/781 [00:21<00:19, 19.34it/s]
 51%|#####1    | 402/781 [00:21<00:19, 19.35it/s]
 52%|#####1    | 404/781 [00:21<00:19, 19.35it/s]
 52%|#####1    | 406/781 [00:21<00:19, 19.26it/s]
 52%|#####2    | 408/781 [00:21<00:19, 19.12it/s]
 52%|#####2    | 410/781 [00:21<00:19, 19.24it/s]
 53%|#####2    | 412/781 [00:21<00:19, 19.08it/s]
 53%|#####3    | 414/781 [00:21<00:19, 19.20it/s]
 53%|#####3    | 416/781 [00:21<00:18, 19.26it/s]
 54%|#####3    | 418/781 [00:22<00:18, 19.23it/s]
 54%|#####3    | 420/781 [00:22<00:18, 19.29it/s]
 54%|#####4    | 422/781 [00:22<00:18, 19.34it/s]
 54%|#####4    | 424/781 [00:22<00:18, 19.32it/s]
 55%|#####4    | 426/781 [00:22<00:18, 19.28it/s]
 55%|#####4    | 428/781 [00:22<00:18, 19.21it/s]
 55%|#####5    | 430/781 [00:22<00:18, 19.10it/s]
 55%|#####5    | 432/781 [00:22<00:18, 18.96it/s]
 56%|#####5    | 434/781 [00:22<00:18, 18.95it/s]
 56%|#####5    | 436/781 [00:23<00:18, 19.02it/s]
 56%|#####6    | 438/781 [00:23<00:18, 19.03it/s]
 56%|#####6    | 440/781 [00:23<00:17, 19.01it/s]
 57%|#####6    | 442/781 [00:23<00:17, 19.12it/s]
 57%|#####6    | 444/781 [00:23<00:17, 19.03it/s]
 57%|#####7    | 446/781 [00:23<00:17, 18.70it/s]
 57%|#####7    | 448/781 [00:23<00:17, 18.60it/s]
 58%|#####7    | 450/781 [00:23<00:17, 18.57it/s]
 58%|#####7    | 452/781 [00:23<00:17, 18.70it/s]
 58%|#####8    | 454/781 [00:23<00:17, 18.68it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.70it/s]
 59%|#####8    | 458/781 [00:24<00:17, 18.79it/s]
 59%|#####8    | 460/781 [00:24<00:16, 18.90it/s]
 59%|#####9    | 462/781 [00:24<00:16, 19.06it/s]
 59%|#####9    | 464/781 [00:24<00:16, 19.08it/s]
 60%|#####9    | 466/781 [00:24<00:16, 18.83it/s]
 60%|#####9    | 468/781 [00:24<00:16, 18.56it/s]
 60%|######    | 470/781 [00:24<00:17, 18.29it/s]
 60%|######    | 472/781 [00:24<00:16, 18.38it/s]
 61%|######    | 474/781 [00:25<00:16, 18.14it/s]
 61%|######    | 476/781 [00:25<00:16, 18.44it/s]
 61%|######1   | 478/781 [00:25<00:16, 18.66it/s]
 61%|######1   | 480/781 [00:25<00:16, 18.70it/s]
 62%|######1   | 482/781 [00:25<00:15, 18.79it/s]
 62%|######1   | 484/781 [00:25<00:15, 18.83it/s]
 62%|######2   | 486/781 [00:25<00:16, 18.43it/s]
 62%|######2   | 488/781 [00:25<00:15, 18.62it/s]
 63%|######2   | 490/781 [00:25<00:15, 18.80it/s]
 63%|######2   | 492/781 [00:26<00:15, 18.68it/s]
 63%|######3   | 494/781 [00:26<00:15, 18.81it/s]
 64%|######3   | 496/781 [00:26<00:15, 18.86it/s]
 64%|######3   | 498/781 [00:26<00:14, 18.91it/s]
 64%|######4   | 500/781 [00:26<00:14, 18.94it/s]
 64%|######4   | 502/781 [00:26<00:14, 19.05it/s]
 65%|######4   | 504/781 [00:26<00:14, 18.79it/s]
 65%|######4   | 506/781 [00:26<00:14, 18.72it/s]
 65%|######5   | 508/781 [00:26<00:14, 18.93it/s]
 65%|######5   | 510/781 [00:26<00:14, 19.06it/s]
 66%|######5   | 512/781 [00:27<00:14, 19.21it/s]
 66%|######5   | 514/781 [00:27<00:13, 19.28it/s]
 66%|######6   | 516/781 [00:27<00:13, 19.19it/s]
 66%|######6   | 518/781 [00:27<00:13, 19.32it/s]
 67%|######6   | 520/781 [00:27<00:13, 19.38it/s]
 67%|######6   | 522/781 [00:27<00:13, 19.38it/s]
 67%|######7   | 524/781 [00:27<00:13, 19.47it/s]
 67%|######7   | 526/781 [00:27<00:13, 19.47it/s]
 68%|######7   | 528/781 [00:27<00:13, 19.45it/s]
 68%|######7   | 530/781 [00:28<00:12, 19.46it/s]
 68%|######8   | 532/781 [00:28<00:12, 19.50it/s]
 68%|######8   | 534/781 [00:28<00:12, 19.52it/s]
 69%|######8   | 536/781 [00:28<00:12, 19.48it/s]
 69%|######8   | 538/781 [00:28<00:12, 19.39it/s]
 69%|######9   | 540/781 [00:28<00:12, 19.22it/s]
 69%|######9   | 542/781 [00:28<00:12, 19.24it/s]
 70%|######9   | 544/781 [00:28<00:12, 19.29it/s]
 70%|######9   | 546/781 [00:28<00:12, 19.27it/s]
 70%|#######   | 548/781 [00:28<00:12, 19.30it/s]
 70%|#######   | 550/781 [00:29<00:11, 19.32it/s]
 71%|#######   | 552/781 [00:29<00:11, 19.24it/s]
 71%|#######   | 554/781 [00:29<00:11, 19.27it/s]
 71%|#######1  | 556/781 [00:29<00:11, 19.24it/s]
 71%|#######1  | 558/781 [00:29<00:11, 19.29it/s]
 72%|#######1  | 560/781 [00:29<00:11, 19.35it/s]
 72%|#######1  | 562/781 [00:29<00:11, 19.43it/s]
 72%|#######2  | 564/781 [00:29<00:11, 19.19it/s]
 72%|#######2  | 566/781 [00:29<00:11, 19.28it/s]
 73%|#######2  | 568/781 [00:29<00:11, 19.21it/s]
 73%|#######2  | 570/781 [00:30<00:11, 19.14it/s]
 73%|#######3  | 572/781 [00:30<00:11, 18.98it/s]
 73%|#######3  | 574/781 [00:30<00:10, 19.06it/s]
 74%|#######3  | 576/781 [00:30<00:10, 18.97it/s]
 74%|#######4  | 578/781 [00:30<00:10, 18.86it/s]
 74%|#######4  | 580/781 [00:30<00:10, 18.87it/s]
 75%|#######4  | 582/781 [00:30<00:10, 19.03it/s]
 75%|#######4  | 584/781 [00:30<00:10, 19.09it/s]
 75%|#######5  | 586/781 [00:30<00:10, 19.17it/s]
 75%|#######5  | 588/781 [00:31<00:10, 19.18it/s]
 76%|#######5  | 590/781 [00:31<00:09, 19.12it/s]
 76%|#######5  | 592/781 [00:31<00:09, 19.15it/s]
 76%|#######6  | 594/781 [00:31<00:09, 19.16it/s]
 76%|#######6  | 596/781 [00:31<00:09, 19.27it/s]
 77%|#######6  | 598/781 [00:31<00:09, 19.31it/s]
 77%|#######6  | 600/781 [00:31<00:09, 19.33it/s]
 77%|#######7  | 602/781 [00:31<00:09, 19.33it/s]
 77%|#######7  | 604/781 [00:31<00:09, 19.28it/s]
 78%|#######7  | 606/781 [00:31<00:09, 19.41it/s]
 78%|#######7  | 608/781 [00:32<00:08, 19.35it/s]
 78%|#######8  | 610/781 [00:32<00:08, 19.32it/s]
 78%|#######8  | 612/781 [00:32<00:08, 19.30it/s]
 79%|#######8  | 614/781 [00:32<00:08, 19.22it/s]
 79%|#######8  | 616/781 [00:32<00:08, 19.00it/s]
 79%|#######9  | 618/781 [00:32<00:08, 19.10it/s]
 79%|#######9  | 620/781 [00:32<00:08, 19.21it/s]
 80%|#######9  | 622/781 [00:32<00:08, 19.16it/s]
 80%|#######9  | 624/781 [00:32<00:08, 19.15it/s]
 80%|########  | 626/781 [00:32<00:08, 19.27it/s]
 80%|########  | 628/781 [00:33<00:07, 19.33it/s]
 81%|########  | 630/781 [00:33<00:07, 19.34it/s]
 81%|########  | 632/781 [00:33<00:07, 19.31it/s]
 81%|########1 | 634/781 [00:33<00:07, 19.37it/s]
 81%|########1 | 636/781 [00:33<00:07, 19.22it/s]
 82%|########1 | 638/781 [00:33<00:07, 19.05it/s]
 82%|########1 | 640/781 [00:33<00:07, 18.90it/s]
 82%|########2 | 642/781 [00:33<00:07, 19.03it/s]
 82%|########2 | 644/781 [00:33<00:07, 19.22it/s]
 83%|########2 | 646/781 [00:34<00:07, 18.96it/s]
 83%|########2 | 648/781 [00:34<00:06, 19.07it/s]
 83%|########3 | 650/781 [00:34<00:06, 19.01it/s]
 83%|########3 | 652/781 [00:34<00:06, 18.84it/s]
 84%|########3 | 654/781 [00:34<00:06, 18.87it/s]
 84%|########3 | 656/781 [00:34<00:06, 18.59it/s]
 84%|########4 | 658/781 [00:34<00:06, 18.31it/s]
 85%|########4 | 660/781 [00:34<00:06, 18.23it/s]
 85%|########4 | 662/781 [00:34<00:06, 18.52it/s]
 85%|########5 | 664/781 [00:35<00:06, 18.60it/s]
 85%|########5 | 666/781 [00:35<00:06, 18.02it/s]
 86%|########5 | 668/781 [00:35<00:06, 18.22it/s]
 86%|########5 | 670/781 [00:35<00:06, 18.30it/s]
 86%|########6 | 672/781 [00:35<00:05, 18.33it/s]
 86%|########6 | 674/781 [00:35<00:05, 18.31it/s]
 87%|########6 | 676/781 [00:35<00:05, 18.33it/s]
 87%|########6 | 678/781 [00:35<00:05, 18.30it/s]
 87%|########7 | 680/781 [00:35<00:05, 18.41it/s]
 87%|########7 | 682/781 [00:35<00:05, 18.59it/s]
 88%|########7 | 684/781 [00:36<00:05, 18.79it/s]
 88%|########7 | 686/781 [00:36<00:05, 18.90it/s]
 88%|########8 | 688/781 [00:36<00:04, 19.01it/s]
 88%|########8 | 690/781 [00:36<00:04, 19.09it/s]
 89%|########8 | 692/781 [00:36<00:04, 19.14it/s]
 89%|########8 | 694/781 [00:36<00:04, 19.10it/s]
 89%|########9 | 696/781 [00:36<00:04, 19.02it/s]
 89%|########9 | 698/781 [00:36<00:04, 19.19it/s]
 90%|########9 | 700/781 [00:36<00:04, 19.29it/s]
 90%|########9 | 702/781 [00:37<00:04, 19.30it/s]
 90%|######### | 704/781 [00:37<00:04, 19.15it/s]
 90%|######### | 706/781 [00:37<00:03, 19.10it/s]
 91%|######### | 708/781 [00:37<00:03, 19.21it/s]
 91%|######### | 710/781 [00:37<00:03, 19.14it/s]
 91%|#########1| 712/781 [00:37<00:03, 19.01it/s]
 91%|#########1| 714/781 [00:37<00:03, 19.03it/s]
 92%|#########1| 716/781 [00:37<00:03, 19.18it/s]
 92%|#########1| 718/781 [00:37<00:03, 18.83it/s]
 92%|#########2| 720/781 [00:37<00:03, 18.94it/s]
 92%|#########2| 722/781 [00:38<00:03, 18.95it/s]
 93%|#########2| 724/781 [00:38<00:02, 19.03it/s]
 93%|#########2| 726/781 [00:38<00:02, 19.14it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.30it/s]
 93%|#########3| 730/781 [00:38<00:02, 19.35it/s]
 94%|#########3| 732/781 [00:38<00:02, 19.40it/s]
 94%|#########3| 734/781 [00:38<00:02, 19.38it/s]
 94%|#########4| 736/781 [00:38<00:02, 19.41it/s]
 94%|#########4| 738/781 [00:38<00:02, 19.45it/s]
 95%|#########4| 740/781 [00:39<00:02, 19.28it/s]
 95%|#########5| 742/781 [00:39<00:02, 18.93it/s]
 95%|#########5| 744/781 [00:39<00:01, 19.04it/s]
 96%|#########5| 746/781 [00:39<00:01, 19.08it/s]
 96%|#########5| 748/781 [00:39<00:01, 19.13it/s]
 96%|#########6| 750/781 [00:39<00:01, 18.99it/s]
 96%|#########6| 752/781 [00:39<00:01, 18.97it/s]
 97%|#########6| 754/781 [00:39<00:01, 18.96it/s]
 97%|#########6| 756/781 [00:39<00:01, 18.97it/s]
 97%|#########7| 758/781 [00:39<00:01, 19.15it/s]
 97%|#########7| 760/781 [00:40<00:01, 19.28it/s]
 98%|#########7| 762/781 [00:40<00:00, 19.23it/s]
 98%|#########7| 764/781 [00:40<00:00, 19.21it/s]
 98%|#########8| 766/781 [00:40<00:00, 19.23it/s]
 98%|#########8| 768/781 [00:40<00:00, 19.13it/s]
 99%|#########8| 770/781 [00:40<00:00, 19.04it/s]
 99%|#########8| 772/781 [00:40<00:00, 19.02it/s]
 99%|#########9| 774/781 [00:40<00:00, 19.09it/s]
 99%|#########9| 776/781 [00:40<00:00, 19.23it/s]
100%|#########9| 778/781 [00:41<00:00, 18.98it/s]
100%|#########9| 780/781 [00:41<00:00, 19.09it/s]
100%|##########| 781/781 [00:41<00:00, 18.97it/s]








Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.16s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.17s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.17s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.18s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.17s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:50,  1.18s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.13s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.23s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.26s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:46,  1.19s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.20s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.21s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.31s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.34s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:04,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.34s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.13s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:05<00:00,  5.60s/it]
100%|##########| 1/1 [00:05<00:00,  5.60s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:09, 16.19it/s]
  3%|2         | 4/157 [00:00<00:09, 15.43it/s]
  4%|3         | 6/157 [00:00<00:09, 16.21it/s]
  5%|5         | 8/157 [00:00<00:08, 16.90it/s]
  6%|6         | 10/157 [00:00<00:08, 17.29it/s]
  8%|7         | 12/157 [00:00<00:08, 17.52it/s]
  9%|8         | 14/157 [00:00<00:08, 17.44it/s]
 10%|#         | 16/157 [00:00<00:07, 17.66it/s]
 11%|#1        | 18/157 [00:01<00:07, 17.71it/s]
 13%|#2        | 20/157 [00:01<00:07, 17.57it/s]
 14%|#4        | 22/157 [00:01<00:07, 17.54it/s]
 15%|#5        | 24/157 [00:01<00:07, 17.59it/s]
 17%|#6        | 26/157 [00:01<00:07, 17.66it/s]
 18%|#7        | 28/157 [00:01<00:07, 17.72it/s]
 19%|#9        | 30/157 [00:01<00:07, 17.86it/s]
 20%|##        | 32/157 [00:01<00:06, 17.94it/s]
 22%|##1       | 34/157 [00:01<00:06, 18.02it/s]
 23%|##2       | 36/157 [00:02<00:06, 18.07it/s]
 24%|##4       | 38/157 [00:02<00:06, 18.11it/s]
 25%|##5       | 40/157 [00:02<00:06, 17.99it/s]
 27%|##6       | 42/157 [00:02<00:06, 17.88it/s]
 28%|##8       | 44/157 [00:02<00:06, 17.85it/s]
 29%|##9       | 46/157 [00:02<00:06, 17.95it/s]
 31%|###       | 48/157 [00:02<00:06, 17.97it/s]
 32%|###1      | 50/157 [00:02<00:05, 18.05it/s]
 33%|###3      | 52/157 [00:02<00:05, 18.13it/s]
 34%|###4      | 54/157 [00:03<00:05, 18.02it/s]
 36%|###5      | 56/157 [00:03<00:05, 17.84it/s]
 37%|###6      | 58/157 [00:03<00:05, 17.89it/s]
 38%|###8      | 60/157 [00:03<00:05, 17.93it/s]
 39%|###9      | 62/157 [00:03<00:05, 17.68it/s]
 41%|####      | 64/157 [00:03<00:05, 17.66it/s]
 42%|####2     | 66/157 [00:03<00:05, 17.57it/s]
 43%|####3     | 68/157 [00:03<00:05, 17.74it/s]
 45%|####4     | 70/157 [00:03<00:04, 17.79it/s]
 46%|####5     | 72/157 [00:04<00:04, 17.93it/s]
 47%|####7     | 74/157 [00:04<00:04, 17.77it/s]
 48%|####8     | 76/157 [00:04<00:04, 17.69it/s]
 50%|####9     | 78/157 [00:04<00:04, 17.64it/s]
 51%|#####     | 80/157 [00:04<00:04, 17.79it/s]
 52%|#####2    | 82/157 [00:04<00:04, 17.94it/s]
 54%|#####3    | 84/157 [00:04<00:04, 17.94it/s]
 55%|#####4    | 86/157 [00:04<00:03, 17.89it/s]
 56%|#####6    | 88/157 [00:04<00:03, 17.80it/s]
 57%|#####7    | 90/157 [00:05<00:03, 17.88it/s]
 59%|#####8    | 92/157 [00:05<00:03, 17.99it/s]
 60%|#####9    | 94/157 [00:05<00:03, 18.08it/s]
 61%|######1   | 96/157 [00:05<00:03, 18.06it/s]
 62%|######2   | 98/157 [00:05<00:03, 18.03it/s]
 64%|######3   | 100/157 [00:05<00:03, 17.87it/s]
 65%|######4   | 102/157 [00:05<00:03, 17.75it/s]
 66%|######6   | 104/157 [00:05<00:02, 17.82it/s]
 68%|######7   | 106/157 [00:05<00:02, 17.86it/s]
 69%|######8   | 108/157 [00:06<00:02, 17.90it/s]
 70%|#######   | 110/157 [00:06<00:02, 17.43it/s]
 71%|#######1  | 112/157 [00:06<00:02, 17.28it/s]
 73%|#######2  | 114/157 [00:06<00:02, 17.51it/s]
 74%|#######3  | 116/157 [00:06<00:02, 17.61it/s]
 75%|#######5  | 118/157 [00:06<00:02, 17.49it/s]
 76%|#######6  | 120/157 [00:06<00:02, 17.52it/s]
 78%|#######7  | 122/157 [00:06<00:01, 17.69it/s]
 79%|#######8  | 124/157 [00:07<00:01, 17.40it/s]
 80%|########  | 126/157 [00:07<00:01, 17.52it/s]
 82%|########1 | 128/157 [00:07<00:01, 17.74it/s]
 83%|########2 | 130/157 [00:07<00:01, 17.90it/s]
 84%|########4 | 132/157 [00:07<00:01, 17.94it/s]
 85%|########5 | 134/157 [00:07<00:01, 17.94it/s]
 87%|########6 | 136/157 [00:07<00:01, 17.96it/s]
 88%|########7 | 138/157 [00:07<00:01, 18.05it/s]
 89%|########9 | 140/157 [00:07<00:00, 17.63it/s]
 90%|######### | 142/157 [00:08<00:00, 17.10it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.94it/s]
 93%|#########2| 146/157 [00:08<00:00, 16.55it/s]
 94%|#########4| 148/157 [00:08<00:00, 16.24it/s]
 96%|#########5| 150/157 [00:08<00:00, 15.45it/s]
 97%|#########6| 152/157 [00:08<00:00, 15.50it/s]
 98%|#########8| 154/157 [00:08<00:00, 15.58it/s]
 99%|#########9| 156/157 [00:08<00:00, 15.67it/s]
100%|##########| 157/157 [00:08<00:00, 17.54it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:51, 15.06it/s]
  1%|          | 4/781 [00:00<00:47, 16.51it/s]
  1%|          | 6/781 [00:00<00:46, 16.76it/s]
  1%|1         | 8/781 [00:00<00:45, 17.05it/s]
  1%|1         | 10/781 [00:00<00:44, 17.42it/s]
  2%|1         | 12/781 [00:00<00:43, 17.63it/s]
  2%|1         | 14/781 [00:00<00:43, 17.78it/s]
  2%|2         | 16/781 [00:00<00:42, 17.83it/s]
  2%|2         | 18/781 [00:01<00:42, 17.77it/s]
  3%|2         | 20/781 [00:01<00:42, 17.70it/s]
  3%|2         | 22/781 [00:01<00:43, 17.60it/s]
  3%|3         | 24/781 [00:01<00:42, 17.65it/s]
  3%|3         | 26/781 [00:01<00:42, 17.67it/s]
  4%|3         | 28/781 [00:01<00:42, 17.80it/s]
  4%|3         | 30/781 [00:01<00:41, 17.93it/s]
  4%|4         | 32/781 [00:01<00:41, 17.95it/s]
  4%|4         | 34/781 [00:01<00:42, 17.72it/s]
  5%|4         | 36/781 [00:02<00:41, 17.79it/s]
  5%|4         | 38/781 [00:02<00:41, 17.90it/s]
  5%|5         | 40/781 [00:02<00:41, 17.99it/s]
  5%|5         | 42/781 [00:02<00:41, 17.82it/s]
  6%|5         | 44/781 [00:02<00:41, 17.81it/s]
  6%|5         | 46/781 [00:02<00:41, 17.85it/s]
  6%|6         | 48/781 [00:02<00:40, 17.97it/s]
  6%|6         | 50/781 [00:02<00:40, 18.07it/s]
  7%|6         | 52/781 [00:02<00:40, 18.14it/s]
  7%|6         | 54/781 [00:03<00:41, 17.64it/s]
  7%|7         | 56/781 [00:03<00:41, 17.60it/s]
  7%|7         | 58/781 [00:03<00:40, 17.65it/s]
  8%|7         | 60/781 [00:03<00:40, 17.69it/s]
  8%|7         | 62/781 [00:03<00:40, 17.61it/s]
  8%|8         | 64/781 [00:03<00:41, 17.39it/s]
  8%|8         | 66/781 [00:03<00:41, 17.26it/s]
  9%|8         | 68/781 [00:03<00:40, 17.56it/s]
  9%|8         | 70/781 [00:03<00:40, 17.54it/s]
  9%|9         | 72/781 [00:04<00:40, 17.62it/s]
  9%|9         | 74/781 [00:04<00:39, 17.70it/s]
 10%|9         | 76/781 [00:04<00:39, 17.74it/s]
 10%|9         | 78/781 [00:04<00:39, 17.72it/s]
 10%|#         | 80/781 [00:04<00:39, 17.88it/s]
 10%|#         | 82/781 [00:04<00:38, 17.99it/s]
 11%|#         | 84/781 [00:04<00:38, 18.01it/s]
 11%|#1        | 86/781 [00:04<00:38, 18.02it/s]
 11%|#1        | 88/781 [00:04<00:38, 17.97it/s]
 12%|#1        | 90/781 [00:05<00:38, 17.79it/s]
 12%|#1        | 92/781 [00:05<00:38, 17.88it/s]
 12%|#2        | 94/781 [00:05<00:38, 17.99it/s]
 12%|#2        | 96/781 [00:05<00:38, 18.01it/s]
 13%|#2        | 98/781 [00:05<00:37, 18.02it/s]
 13%|#2        | 100/781 [00:05<00:37, 18.07it/s]
 13%|#3        | 102/781 [00:05<00:37, 17.97it/s]
 13%|#3        | 104/781 [00:05<00:37, 17.98it/s]
 14%|#3        | 106/781 [00:05<00:37, 18.05it/s]
 14%|#3        | 108/781 [00:06<00:37, 17.89it/s]
 14%|#4        | 110/781 [00:06<00:37, 17.97it/s]
 14%|#4        | 112/781 [00:06<00:37, 17.84it/s]
 15%|#4        | 114/781 [00:06<00:37, 17.77it/s]
 15%|#4        | 116/781 [00:06<00:37, 17.69it/s]
 15%|#5        | 118/781 [00:06<00:37, 17.80it/s]
 15%|#5        | 120/781 [00:06<00:37, 17.86it/s]
 16%|#5        | 122/781 [00:06<00:36, 17.95it/s]
 16%|#5        | 124/781 [00:06<00:36, 18.02it/s]
 16%|#6        | 126/781 [00:07<00:36, 18.06it/s]
 16%|#6        | 128/781 [00:07<00:36, 18.09it/s]
 17%|#6        | 130/781 [00:07<00:36, 17.96it/s]
 17%|#6        | 132/781 [00:07<00:36, 17.88it/s]
 17%|#7        | 134/781 [00:07<00:36, 17.92it/s]
 17%|#7        | 136/781 [00:07<00:35, 18.00it/s]
 18%|#7        | 138/781 [00:07<00:35, 18.06it/s]
 18%|#7        | 140/781 [00:07<00:35, 17.96it/s]
 18%|#8        | 142/781 [00:07<00:35, 17.90it/s]
 18%|#8        | 144/781 [00:08<00:35, 17.86it/s]
 19%|#8        | 146/781 [00:08<00:35, 17.74it/s]
 19%|#8        | 148/781 [00:08<00:35, 17.76it/s]
 19%|#9        | 150/781 [00:08<00:35, 17.91it/s]
 19%|#9        | 152/781 [00:08<00:35, 17.70it/s]
 20%|#9        | 154/781 [00:08<00:35, 17.78it/s]
 20%|#9        | 156/781 [00:08<00:35, 17.42it/s]
 20%|##        | 158/781 [00:08<00:35, 17.43it/s]
 20%|##        | 160/781 [00:09<00:35, 17.70it/s]
 21%|##        | 162/781 [00:09<00:33, 18.23it/s]
 21%|##        | 164/781 [00:09<00:33, 18.26it/s]
 21%|##1       | 166/781 [00:09<00:32, 18.68it/s]
 22%|##1       | 168/781 [00:09<00:33, 18.22it/s]
 22%|##1       | 170/781 [00:09<00:33, 18.45it/s]
 22%|##2       | 173/781 [00:09<00:31, 19.06it/s]
 22%|##2       | 175/781 [00:09<00:31, 19.12it/s]
 23%|##2       | 178/781 [00:09<00:31, 19.30it/s]
 23%|##3       | 180/781 [00:10<00:31, 19.37it/s]
 23%|##3       | 182/781 [00:10<00:31, 18.85it/s]
 24%|##3       | 184/781 [00:10<00:32, 18.54it/s]
 24%|##3       | 186/781 [00:10<00:31, 18.83it/s]
 24%|##4       | 188/781 [00:10<00:31, 18.74it/s]
 24%|##4       | 190/781 [00:10<00:31, 18.66it/s]
 25%|##4       | 192/781 [00:10<00:31, 18.84it/s]
 25%|##4       | 195/781 [00:10<00:30, 19.13it/s]
 25%|##5       | 197/781 [00:10<00:30, 19.35it/s]
 26%|##5       | 200/781 [00:11<00:29, 19.68it/s]
 26%|##5       | 203/781 [00:11<00:28, 20.00it/s]
 26%|##6       | 206/781 [00:11<00:28, 20.20it/s]
 27%|##6       | 209/781 [00:11<00:28, 20.34it/s]
 27%|##7       | 212/781 [00:11<00:27, 20.46it/s]
 28%|##7       | 215/781 [00:11<00:27, 20.58it/s]
 28%|##7       | 218/781 [00:11<00:27, 20.62it/s]
 28%|##8       | 221/781 [00:12<00:27, 20.61it/s]
 29%|##8       | 224/781 [00:12<00:26, 20.66it/s]
 29%|##9       | 227/781 [00:12<00:26, 20.67it/s]
 29%|##9       | 230/781 [00:12<00:26, 20.58it/s]
 30%|##9       | 233/781 [00:12<00:27, 20.12it/s]
 30%|###       | 236/781 [00:12<00:35, 15.49it/s]
 30%|###       | 238/781 [00:13<00:33, 16.29it/s]
 31%|###       | 241/781 [00:13<00:30, 17.51it/s]
 31%|###1      | 244/781 [00:13<00:29, 18.26it/s]
 32%|###1      | 247/781 [00:13<00:28, 18.80it/s]
 32%|###2      | 250/781 [00:13<00:27, 19.31it/s]
 32%|###2      | 253/781 [00:13<00:26, 19.62it/s]
 33%|###2      | 256/781 [00:13<00:26, 19.86it/s]
 33%|###3      | 259/781 [00:14<00:26, 19.92it/s]
 34%|###3      | 262/781 [00:14<00:25, 20.15it/s]
 34%|###3      | 265/781 [00:14<00:25, 20.32it/s]
 34%|###4      | 268/781 [00:14<00:25, 20.46it/s]
 35%|###4      | 271/781 [00:14<00:24, 20.41it/s]
 35%|###5      | 274/781 [00:14<00:24, 20.54it/s]
 35%|###5      | 277/781 [00:15<00:24, 20.47it/s]
 36%|###5      | 280/781 [00:15<00:24, 20.38it/s]
 36%|###6      | 283/781 [00:15<00:24, 20.28it/s]
 37%|###6      | 286/781 [00:15<00:24, 20.13it/s]
 37%|###7      | 289/781 [00:15<00:24, 20.23it/s]
 37%|###7      | 292/781 [00:15<00:24, 19.83it/s]
 38%|###7      | 295/781 [00:15<00:24, 20.03it/s]
 38%|###8      | 298/781 [00:16<00:23, 20.19it/s]
 39%|###8      | 301/781 [00:16<00:23, 20.34it/s]
 39%|###8      | 304/781 [00:16<00:23, 20.48it/s]
 39%|###9      | 307/781 [00:16<00:23, 20.28it/s]
 40%|###9      | 310/781 [00:16<00:23, 20.31it/s]
 40%|####      | 313/781 [00:16<00:23, 20.25it/s]
 40%|####      | 316/781 [00:16<00:23, 20.19it/s]
 41%|####      | 319/781 [00:17<00:22, 20.33it/s]
 41%|####1     | 322/781 [00:17<00:22, 20.53it/s]
 42%|####1     | 325/781 [00:17<00:22, 20.56it/s]
 42%|####1     | 328/781 [00:17<00:22, 20.53it/s]
 42%|####2     | 331/781 [00:17<00:22, 20.16it/s]
 43%|####2     | 334/781 [00:17<00:22, 19.53it/s]
 43%|####3     | 336/781 [00:17<00:22, 19.62it/s]
 43%|####3     | 338/781 [00:18<00:22, 19.70it/s]
 44%|####3     | 340/781 [00:18<00:22, 19.71it/s]
 44%|####3     | 343/781 [00:18<00:21, 20.03it/s]
 44%|####4     | 346/781 [00:18<00:21, 20.31it/s]
 45%|####4     | 349/781 [00:18<00:21, 20.40it/s]
 45%|####5     | 352/781 [00:18<00:20, 20.47it/s]
 45%|####5     | 355/781 [00:18<00:20, 20.50it/s]
 46%|####5     | 358/781 [00:19<00:20, 20.29it/s]
 46%|####6     | 361/781 [00:19<00:20, 20.14it/s]
 47%|####6     | 364/781 [00:19<00:20, 20.13it/s]
 47%|####6     | 367/781 [00:19<00:20, 20.25it/s]
 47%|####7     | 370/781 [00:19<00:20, 20.18it/s]
 48%|####7     | 373/781 [00:19<00:20, 20.17it/s]
 48%|####8     | 376/781 [00:19<00:19, 20.32it/s]
 49%|####8     | 379/781 [00:20<00:19, 20.30it/s]
 49%|####8     | 382/781 [00:20<00:19, 20.41it/s]
 49%|####9     | 385/781 [00:20<00:19, 20.51it/s]
 50%|####9     | 388/781 [00:20<00:19, 20.43it/s]
 50%|#####     | 391/781 [00:20<00:19, 20.41it/s]
 50%|#####     | 394/781 [00:20<00:19, 20.33it/s]
 51%|#####     | 397/781 [00:20<00:18, 20.44it/s]
 51%|#####1    | 400/781 [00:21<00:18, 20.58it/s]
 52%|#####1    | 403/781 [00:21<00:18, 20.66it/s]
 52%|#####1    | 406/781 [00:21<00:18, 20.58it/s]
 52%|#####2    | 409/781 [00:21<00:18, 20.63it/s]
 53%|#####2    | 412/781 [00:21<00:17, 20.52it/s]
 53%|#####3    | 415/781 [00:21<00:17, 20.64it/s]
 54%|#####3    | 418/781 [00:21<00:17, 20.55it/s]
 54%|#####3    | 421/781 [00:22<00:17, 20.48it/s]
 54%|#####4    | 424/781 [00:22<00:17, 20.58it/s]
 55%|#####4    | 427/781 [00:22<00:17, 20.63it/s]
 55%|#####5    | 430/781 [00:22<00:16, 20.66it/s]
 55%|#####5    | 433/781 [00:22<00:16, 20.63it/s]
 56%|#####5    | 436/781 [00:22<00:16, 20.57it/s]
 56%|#####6    | 439/781 [00:22<00:16, 20.44it/s]
 57%|#####6    | 442/781 [00:23<00:16, 20.51it/s]
 57%|#####6    | 445/781 [00:23<00:16, 20.36it/s]
 57%|#####7    | 448/781 [00:23<00:16, 20.48it/s]
 58%|#####7    | 451/781 [00:23<00:16, 20.53it/s]
 58%|#####8    | 454/781 [00:23<00:15, 20.54it/s]
 59%|#####8    | 457/781 [00:23<00:15, 20.60it/s]
 59%|#####8    | 460/781 [00:23<00:15, 20.61it/s]
 59%|#####9    | 463/781 [00:24<00:15, 20.68it/s]
 60%|#####9    | 466/781 [00:24<00:15, 20.66it/s]
 60%|######    | 469/781 [00:24<00:15, 20.50it/s]
 60%|######    | 472/781 [00:24<00:15, 20.40it/s]
 61%|######    | 475/781 [00:24<00:14, 20.53it/s]
 61%|######1   | 478/781 [00:24<00:14, 20.65it/s]
 62%|######1   | 481/781 [00:25<00:14, 20.63it/s]
 62%|######1   | 484/781 [00:25<00:14, 20.50it/s]
 62%|######2   | 487/781 [00:25<00:14, 20.57it/s]
 63%|######2   | 490/781 [00:25<00:14, 20.56it/s]
 63%|######3   | 493/781 [00:25<00:14, 20.44it/s]
 64%|######3   | 496/781 [00:25<00:13, 20.51it/s]
 64%|######3   | 499/781 [00:25<00:13, 20.37it/s]
 64%|######4   | 502/781 [00:26<00:13, 20.23it/s]
 65%|######4   | 505/781 [00:26<00:13, 20.22it/s]
 65%|######5   | 508/781 [00:26<00:13, 20.26it/s]
 65%|######5   | 511/781 [00:26<00:13, 20.30it/s]
 66%|######5   | 514/781 [00:26<00:13, 20.40it/s]
 66%|######6   | 517/781 [00:26<00:12, 20.49it/s]
 67%|######6   | 520/781 [00:26<00:12, 20.50it/s]
 67%|######6   | 523/781 [00:27<00:12, 20.40it/s]
 67%|######7   | 526/781 [00:27<00:12, 20.36it/s]
 68%|######7   | 529/781 [00:27<00:12, 20.21it/s]
 68%|######8   | 532/781 [00:27<00:12, 20.41it/s]
 69%|######8   | 535/781 [00:27<00:11, 20.52it/s]
 69%|######8   | 538/781 [00:27<00:11, 20.44it/s]
 69%|######9   | 541/781 [00:27<00:11, 20.61it/s]
 70%|######9   | 544/781 [00:28<00:11, 20.68it/s]
 70%|#######   | 547/781 [00:28<00:11, 20.68it/s]
 70%|#######   | 550/781 [00:28<00:11, 20.52it/s]
 71%|#######   | 553/781 [00:28<00:11, 20.51it/s]
 71%|#######1  | 556/781 [00:28<00:11, 20.44it/s]
 72%|#######1  | 559/781 [00:28<00:11, 20.09it/s]
 72%|#######1  | 562/781 [00:28<00:10, 20.12it/s]
 72%|#######2  | 565/781 [00:29<00:10, 19.90it/s]
 73%|#######2  | 568/781 [00:29<00:10, 19.96it/s]
 73%|#######2  | 570/781 [00:29<00:10, 19.73it/s]
 73%|#######3  | 572/781 [00:29<00:10, 19.59it/s]
 73%|#######3  | 574/781 [00:29<00:10, 19.68it/s]
 74%|#######3  | 577/781 [00:29<00:10, 19.87it/s]
 74%|#######4  | 580/781 [00:29<00:10, 20.04it/s]
 75%|#######4  | 583/781 [00:30<00:09, 20.11it/s]
 75%|#######5  | 586/781 [00:30<00:09, 20.20it/s]
 75%|#######5  | 589/781 [00:30<00:09, 20.30it/s]
 76%|#######5  | 592/781 [00:30<00:09, 20.08it/s]
 76%|#######6  | 595/781 [00:30<00:09, 20.30it/s]
 77%|#######6  | 598/781 [00:30<00:09, 20.33it/s]
 77%|#######6  | 601/781 [00:30<00:08, 20.16it/s]
 77%|#######7  | 604/781 [00:31<00:08, 20.09it/s]
 78%|#######7  | 607/781 [00:31<00:08, 20.31it/s]
 78%|#######8  | 610/781 [00:31<00:08, 20.47it/s]
 78%|#######8  | 613/781 [00:31<00:08, 20.49it/s]
 79%|#######8  | 616/781 [00:31<00:08, 20.53it/s]
 79%|#######9  | 619/781 [00:31<00:07, 20.64it/s]
 80%|#######9  | 622/781 [00:31<00:07, 20.76it/s]
 80%|########  | 625/781 [00:32<00:07, 20.78it/s]
 80%|########  | 628/781 [00:32<00:07, 20.75it/s]
 81%|########  | 631/781 [00:32<00:07, 20.74it/s]
 81%|########1 | 634/781 [00:32<00:07, 20.79it/s]
 82%|########1 | 637/781 [00:32<00:06, 20.61it/s]
 82%|########1 | 640/781 [00:32<00:06, 20.28it/s]
 82%|########2 | 643/781 [00:32<00:06, 20.31it/s]
 83%|########2 | 646/781 [00:33<00:06, 20.48it/s]
 83%|########3 | 649/781 [00:33<00:06, 20.60it/s]
 83%|########3 | 652/781 [00:33<00:06, 20.67it/s]
 84%|########3 | 655/781 [00:33<00:06, 20.56it/s]
 84%|########4 | 658/781 [00:33<00:05, 20.57it/s]
 85%|########4 | 661/781 [00:33<00:05, 20.51it/s]
 85%|########5 | 664/781 [00:33<00:05, 20.54it/s]
 85%|########5 | 667/781 [00:34<00:05, 20.58it/s]
 86%|########5 | 670/781 [00:34<00:05, 20.69it/s]
 86%|########6 | 673/781 [00:34<00:05, 20.58it/s]
 87%|########6 | 676/781 [00:34<00:05, 20.31it/s]
 87%|########6 | 679/781 [00:34<00:05, 20.24it/s]
 87%|########7 | 682/781 [00:34<00:04, 20.41it/s]
 88%|########7 | 685/781 [00:35<00:04, 20.46it/s]
 88%|########8 | 688/781 [00:35<00:04, 20.24it/s]
 88%|########8 | 691/781 [00:35<00:04, 20.22it/s]
 89%|########8 | 694/781 [00:35<00:04, 20.23it/s]
 89%|########9 | 697/781 [00:35<00:04, 20.33it/s]
 90%|########9 | 700/781 [00:35<00:03, 20.37it/s]
 90%|######### | 703/781 [00:35<00:03, 20.08it/s]
 90%|######### | 706/781 [00:36<00:03, 20.10it/s]
 91%|######### | 709/781 [00:36<00:03, 20.15it/s]
 91%|#########1| 712/781 [00:36<00:03, 20.16it/s]
 92%|#########1| 715/781 [00:36<00:03, 20.30it/s]
 92%|#########1| 718/781 [00:36<00:03, 20.39it/s]
 92%|#########2| 721/781 [00:36<00:02, 20.52it/s]
 93%|#########2| 724/781 [00:36<00:02, 20.45it/s]
 93%|#########3| 727/781 [00:37<00:02, 20.40it/s]
 93%|#########3| 730/781 [00:37<00:02, 20.28it/s]
 94%|#########3| 733/781 [00:37<00:02, 20.43it/s]
 94%|#########4| 736/781 [00:37<00:02, 20.46it/s]
 95%|#########4| 739/781 [00:37<00:02, 20.61it/s]
 95%|#########5| 742/781 [00:37<00:01, 20.54it/s]
 95%|#########5| 745/781 [00:37<00:01, 20.54it/s]
 96%|#########5| 748/781 [00:38<00:01, 20.63it/s]
 96%|#########6| 751/781 [00:38<00:01, 20.47it/s]
 97%|#########6| 754/781 [00:38<00:01, 20.57it/s]
 97%|#########6| 757/781 [00:38<00:01, 20.64it/s]
 97%|#########7| 760/781 [00:38<00:01, 20.72it/s]
 98%|#########7| 763/781 [00:38<00:00, 20.61it/s]
 98%|#########8| 766/781 [00:38<00:00, 20.47it/s]
 98%|#########8| 769/781 [00:39<00:00, 20.48it/s]
 99%|#########8| 772/781 [00:39<00:00, 20.41it/s]
 99%|#########9| 775/781 [00:39<00:00, 20.33it/s]
100%|#########9| 778/781 [00:39<00:00, 20.45it/s]
100%|##########| 781/781 [00:39<00:00, 20.38it/s]
100%|##########| 781/781 [00:39<00:00, 19.66it/s]








Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.13s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:48,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.14s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:49,  1.15s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:52,  1.21s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:46,  1.11s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:46,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:46,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.12s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:47,  1.14s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:49,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:49,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:49,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:50,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:47,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.20s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.16s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.16s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:46,  1.19s/it]
Validation Progress:  14%|#3        | 6/44 [00:06<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:06<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:06<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:06<00:43,  1.15s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:06<00:43,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:43,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:42,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:42,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:42,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:42,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:42,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.18s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:41,  1.19s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.22s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:43,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.21s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.21s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:38,  1.21s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:37,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:33,  1.27s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.28s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.30s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:22,  1.31s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.29s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:20,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:16,  1.35s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.32s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:52<00:03,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.11s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.12s/it]
Validation Progress: 100%|##########| 44/44 [00:55<00:00,  1.26s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:06<00:00,  6.45s/it]
100%|##########| 1/1 [00:06<00:00,  6.45s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.52it/s]
  3%|2         | 4/157 [00:00<00:10, 14.70it/s]
  4%|3         | 6/157 [00:00<00:09, 15.44it/s]
  5%|5         | 8/157 [00:00<00:09, 15.70it/s]
  6%|6         | 10/157 [00:00<00:09, 15.81it/s]
  8%|7         | 12/157 [00:00<00:09, 15.97it/s]
  9%|8         | 14/157 [00:00<00:08, 16.01it/s]
 10%|#         | 16/157 [00:01<00:08, 16.07it/s]
 11%|#1        | 18/157 [00:01<00:08, 15.56it/s]
 13%|#2        | 20/157 [00:01<00:08, 15.72it/s]
 14%|#4        | 22/157 [00:01<00:08, 15.89it/s]
 15%|#5        | 24/157 [00:01<00:08, 16.01it/s]
 17%|#6        | 26/157 [00:01<00:08, 16.13it/s]
 18%|#7        | 28/157 [00:01<00:07, 16.13it/s]
 19%|#9        | 30/157 [00:01<00:07, 16.20it/s]
 20%|##        | 32/157 [00:02<00:07, 16.07it/s]
 22%|##1       | 34/157 [00:02<00:07, 16.09it/s]
 23%|##2       | 36/157 [00:02<00:07, 16.29it/s]
 24%|##4       | 38/157 [00:02<00:07, 16.44it/s]
 25%|##5       | 40/157 [00:02<00:07, 16.56it/s]
 27%|##6       | 42/157 [00:02<00:06, 16.60it/s]
 28%|##8       | 44/157 [00:02<00:06, 16.39it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.30it/s]
 31%|###       | 48/157 [00:02<00:06, 16.34it/s]
 32%|###1      | 50/157 [00:03<00:06, 16.04it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.10it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.08it/s]
 36%|###5      | 56/157 [00:03<00:06, 16.11it/s]
 37%|###6      | 58/157 [00:03<00:06, 16.20it/s]
 38%|###8      | 60/157 [00:03<00:06, 16.09it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.22it/s]
 41%|####      | 64/157 [00:03<00:05, 15.98it/s]
 42%|####2     | 66/157 [00:04<00:05, 16.07it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.15it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.25it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.31it/s]
 47%|####7     | 74/157 [00:04<00:05, 16.32it/s]
 48%|####8     | 76/157 [00:04<00:04, 16.35it/s]
 50%|####9     | 78/157 [00:04<00:04, 16.39it/s]
 51%|#####     | 80/157 [00:04<00:04, 16.35it/s]
 52%|#####2    | 82/157 [00:05<00:04, 16.27it/s]
 54%|#####3    | 84/157 [00:05<00:04, 16.33it/s]
 55%|#####4    | 86/157 [00:05<00:04, 16.32it/s]
 56%|#####6    | 88/157 [00:05<00:04, 16.25it/s]
 57%|#####7    | 90/157 [00:05<00:04, 16.30it/s]
 59%|#####8    | 92/157 [00:05<00:04, 16.13it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.23it/s]
 61%|######1   | 96/157 [00:05<00:03, 16.18it/s]
 62%|######2   | 98/157 [00:06<00:03, 16.04it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.04it/s]
 65%|######4   | 102/157 [00:06<00:03, 15.85it/s]
 66%|######6   | 104/157 [00:06<00:03, 15.97it/s]
 68%|######7   | 106/157 [00:06<00:03, 15.92it/s]
 69%|######8   | 108/157 [00:06<00:03, 15.82it/s]
 70%|#######   | 110/157 [00:06<00:02, 15.72it/s]
 71%|#######1  | 112/157 [00:06<00:02, 15.43it/s]
 73%|#######2  | 114/157 [00:07<00:02, 15.25it/s]
 74%|#######3  | 116/157 [00:07<00:02, 15.31it/s]
 75%|#######5  | 118/157 [00:07<00:02, 15.48it/s]
 76%|#######6  | 120/157 [00:07<00:02, 15.47it/s]
 78%|#######7  | 122/157 [00:07<00:02, 15.48it/s]
 79%|#######8  | 124/157 [00:07<00:02, 15.45it/s]
 80%|########  | 126/157 [00:07<00:02, 15.40it/s]
 82%|########1 | 128/157 [00:08<00:01, 15.43it/s]
 83%|########2 | 130/157 [00:08<00:01, 15.45it/s]
 84%|########4 | 132/157 [00:08<00:01, 15.40it/s]
 85%|########5 | 134/157 [00:08<00:01, 15.42it/s]
 87%|########6 | 136/157 [00:08<00:01, 15.43it/s]
 88%|########7 | 138/157 [00:08<00:01, 15.55it/s]
 89%|########9 | 140/157 [00:08<00:01, 15.67it/s]
 90%|######### | 142/157 [00:08<00:00, 15.56it/s]
 92%|#########1| 144/157 [00:09<00:00, 15.60it/s]
 93%|#########2| 146/157 [00:09<00:00, 15.47it/s]
 94%|#########4| 148/157 [00:09<00:00, 15.51it/s]
 96%|#########5| 150/157 [00:09<00:00, 15.52it/s]
 97%|#########6| 152/157 [00:09<00:00, 15.67it/s]
 98%|#########8| 154/157 [00:09<00:00, 15.92it/s]
 99%|#########9| 156/157 [00:09<00:00, 16.12it/s]
100%|##########| 157/157 [00:09<00:00, 15.97it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:49, 15.61it/s]
  1%|          | 4/781 [00:00<00:45, 17.10it/s]
  1%|          | 6/781 [00:00<00:43, 17.66it/s]
  1%|1         | 8/781 [00:00<00:43, 17.96it/s]
  1%|1         | 10/781 [00:00<00:42, 18.14it/s]
  2%|1         | 12/781 [00:00<00:42, 18.23it/s]
  2%|1         | 14/781 [00:00<00:41, 18.32it/s]
  2%|2         | 16/781 [00:00<00:42, 18.07it/s]
  2%|2         | 18/781 [00:01<00:42, 18.15it/s]
  3%|2         | 20/781 [00:01<00:41, 18.30it/s]
  3%|2         | 22/781 [00:01<00:41, 18.40it/s]
  3%|3         | 24/781 [00:01<00:40, 18.58it/s]
  3%|3         | 26/781 [00:01<00:41, 18.40it/s]
  4%|3         | 28/781 [00:01<00:41, 18.34it/s]
  4%|3         | 30/781 [00:01<00:40, 18.33it/s]
  4%|4         | 32/781 [00:01<00:40, 18.32it/s]
  4%|4         | 34/781 [00:01<00:41, 18.20it/s]
  5%|4         | 36/781 [00:01<00:40, 18.31it/s]
  5%|4         | 38/781 [00:02<00:40, 18.41it/s]
  5%|5         | 40/781 [00:02<00:40, 18.23it/s]
  5%|5         | 42/781 [00:02<00:40, 18.27it/s]
  6%|5         | 44/781 [00:02<00:40, 18.08it/s]
  6%|5         | 46/781 [00:02<00:40, 18.07it/s]
  6%|6         | 48/781 [00:02<00:40, 18.15it/s]
  6%|6         | 50/781 [00:02<00:40, 18.23it/s]
  7%|6         | 52/781 [00:02<00:40, 17.91it/s]
  7%|6         | 54/781 [00:02<00:41, 17.69it/s]
  7%|7         | 56/781 [00:03<00:41, 17.45it/s]
  7%|7         | 58/781 [00:03<00:41, 17.59it/s]
  8%|7         | 60/781 [00:03<00:40, 17.59it/s]
  8%|7         | 62/781 [00:03<00:40, 17.68it/s]
  8%|8         | 64/781 [00:03<00:40, 17.84it/s]
  8%|8         | 66/781 [00:03<00:39, 18.00it/s]
  9%|8         | 68/781 [00:03<00:39, 18.10it/s]
  9%|8         | 70/781 [00:03<00:39, 18.14it/s]
  9%|9         | 72/781 [00:03<00:38, 18.29it/s]
  9%|9         | 74/781 [00:04<00:38, 18.51it/s]
 10%|9         | 76/781 [00:04<00:37, 18.66it/s]
 10%|9         | 78/781 [00:04<00:37, 18.70it/s]
 10%|#         | 80/781 [00:04<00:37, 18.53it/s]
 10%|#         | 82/781 [00:04<00:38, 18.34it/s]
 11%|#         | 84/781 [00:04<00:37, 18.36it/s]
 11%|#1        | 86/781 [00:04<00:37, 18.37it/s]
 11%|#1        | 88/781 [00:04<00:37, 18.29it/s]
 12%|#1        | 90/781 [00:04<00:37, 18.31it/s]
 12%|#1        | 92/781 [00:05<00:37, 18.45it/s]
 12%|#2        | 94/781 [00:05<00:36, 18.63it/s]
 12%|#2        | 96/781 [00:05<00:37, 18.45it/s]
 13%|#2        | 98/781 [00:05<00:36, 18.58it/s]
 13%|#2        | 100/781 [00:05<00:36, 18.47it/s]
 13%|#3        | 102/781 [00:05<00:36, 18.39it/s]
 13%|#3        | 104/781 [00:05<00:37, 18.23it/s]
 14%|#3        | 106/781 [00:05<00:36, 18.28it/s]
 14%|#3        | 108/781 [00:05<00:36, 18.29it/s]
 14%|#4        | 110/781 [00:06<00:36, 18.30it/s]
 14%|#4        | 112/781 [00:06<00:36, 18.34it/s]
 15%|#4        | 114/781 [00:06<00:36, 18.40it/s]
 15%|#4        | 116/781 [00:06<00:35, 18.50it/s]
 15%|#5        | 118/781 [00:06<00:35, 18.59it/s]
 15%|#5        | 120/781 [00:06<00:35, 18.51it/s]
 16%|#5        | 122/781 [00:06<00:35, 18.46it/s]
 16%|#5        | 124/781 [00:06<00:35, 18.37it/s]
 16%|#6        | 126/781 [00:06<00:35, 18.20it/s]
 16%|#6        | 128/781 [00:07<00:35, 18.26it/s]
 17%|#6        | 130/781 [00:07<00:35, 18.33it/s]
 17%|#6        | 132/781 [00:07<00:35, 18.39it/s]
 17%|#7        | 134/781 [00:07<00:35, 18.48it/s]
 17%|#7        | 136/781 [00:07<00:34, 18.45it/s]
 18%|#7        | 138/781 [00:07<00:34, 18.47it/s]
 18%|#7        | 140/781 [00:07<00:34, 18.42it/s]
 18%|#8        | 142/781 [00:07<00:34, 18.28it/s]
 18%|#8        | 144/781 [00:07<00:34, 18.32it/s]
 19%|#8        | 146/781 [00:08<00:34, 18.40it/s]
 19%|#8        | 148/781 [00:08<00:34, 18.52it/s]
 19%|#9        | 150/781 [00:08<00:34, 18.51it/s]
 19%|#9        | 152/781 [00:08<00:33, 18.68it/s]
 20%|#9        | 154/781 [00:08<00:33, 18.70it/s]
 20%|#9        | 156/781 [00:08<00:33, 18.80it/s]
 20%|##        | 158/781 [00:08<00:32, 18.88it/s]
 20%|##        | 160/781 [00:08<00:32, 18.95it/s]
 21%|##        | 162/781 [00:08<00:32, 18.87it/s]
 21%|##        | 164/781 [00:08<00:32, 18.89it/s]
 21%|##1       | 166/781 [00:09<00:32, 18.85it/s]
 22%|##1       | 168/781 [00:09<00:32, 18.91it/s]
 22%|##1       | 170/781 [00:09<00:32, 18.85it/s]
 22%|##2       | 172/781 [00:09<00:32, 18.88it/s]
 22%|##2       | 174/781 [00:09<00:32, 18.80it/s]
 23%|##2       | 176/781 [00:09<00:32, 18.63it/s]
 23%|##2       | 178/781 [00:09<00:32, 18.42it/s]
 23%|##3       | 180/781 [00:09<00:32, 18.43it/s]
 23%|##3       | 182/781 [00:09<00:32, 18.44it/s]
 24%|##3       | 184/781 [00:10<00:32, 18.49it/s]
 24%|##3       | 186/781 [00:10<00:32, 18.50it/s]
 24%|##4       | 188/781 [00:10<00:32, 18.41it/s]
 24%|##4       | 190/781 [00:10<00:32, 18.34it/s]
 25%|##4       | 192/781 [00:10<00:31, 18.41it/s]
 25%|##4       | 194/781 [00:10<00:31, 18.46it/s]
 25%|##5       | 196/781 [00:10<00:32, 18.28it/s]
 25%|##5       | 198/781 [00:10<00:31, 18.42it/s]
 26%|##5       | 200/781 [00:10<00:31, 18.46it/s]
 26%|##5       | 202/781 [00:11<00:31, 18.48it/s]
 26%|##6       | 204/781 [00:11<00:31, 18.48it/s]
 26%|##6       | 206/781 [00:11<00:31, 18.49it/s]
 27%|##6       | 208/781 [00:11<00:30, 18.49it/s]
 27%|##6       | 210/781 [00:11<00:30, 18.49it/s]
 27%|##7       | 212/781 [00:11<00:30, 18.41it/s]
 27%|##7       | 214/781 [00:11<00:30, 18.32it/s]
 28%|##7       | 216/781 [00:11<00:30, 18.32it/s]
 28%|##7       | 218/781 [00:11<00:30, 18.28it/s]
 28%|##8       | 220/781 [00:11<00:30, 18.33it/s]
 28%|##8       | 222/781 [00:12<00:30, 18.46it/s]
 29%|##8       | 224/781 [00:12<00:29, 18.59it/s]
 29%|##8       | 226/781 [00:12<00:30, 18.47it/s]
 29%|##9       | 228/781 [00:12<00:29, 18.51it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.51it/s]
 30%|##9       | 232/781 [00:12<00:29, 18.49it/s]
 30%|##9       | 234/781 [00:12<00:29, 18.50it/s]
 30%|###       | 236/781 [00:12<00:29, 18.32it/s]
 30%|###       | 238/781 [00:12<00:29, 18.27it/s]
 31%|###       | 240/781 [00:13<00:29, 18.27it/s]
 31%|###       | 242/781 [00:13<00:29, 18.33it/s]
 31%|###1      | 244/781 [00:13<00:29, 18.35it/s]
 31%|###1      | 246/781 [00:13<00:29, 18.38it/s]
 32%|###1      | 248/781 [00:13<00:29, 18.37it/s]
 32%|###2      | 250/781 [00:13<00:29, 18.31it/s]
 32%|###2      | 252/781 [00:13<00:28, 18.41it/s]
 33%|###2      | 254/781 [00:13<00:29, 18.11it/s]
 33%|###2      | 256/781 [00:13<00:28, 18.21it/s]
 33%|###3      | 258/781 [00:14<00:28, 18.30it/s]
 33%|###3      | 260/781 [00:14<00:28, 18.13it/s]
 34%|###3      | 262/781 [00:14<00:28, 18.10it/s]
 34%|###3      | 264/781 [00:14<00:28, 18.23it/s]
 34%|###4      | 266/781 [00:14<00:28, 18.37it/s]
 34%|###4      | 268/781 [00:14<00:28, 18.19it/s]
 35%|###4      | 270/781 [00:14<00:27, 18.35it/s]
 35%|###4      | 272/781 [00:14<00:27, 18.45it/s]
 35%|###5      | 274/781 [00:14<00:27, 18.64it/s]
 35%|###5      | 276/781 [00:15<00:26, 18.77it/s]
 36%|###5      | 278/781 [00:15<00:26, 18.87it/s]
 36%|###5      | 280/781 [00:15<00:26, 18.83it/s]
 36%|###6      | 282/781 [00:15<00:26, 18.73it/s]
 36%|###6      | 284/781 [00:15<00:26, 18.75it/s]
 37%|###6      | 286/781 [00:15<00:26, 18.85it/s]
 37%|###6      | 288/781 [00:15<00:26, 18.87it/s]
 37%|###7      | 290/781 [00:15<00:25, 18.96it/s]
 37%|###7      | 292/781 [00:15<00:25, 19.05it/s]
 38%|###7      | 294/781 [00:15<00:25, 18.99it/s]
 38%|###7      | 296/781 [00:16<00:25, 18.85it/s]
 38%|###8      | 298/781 [00:16<00:25, 18.82it/s]
 38%|###8      | 300/781 [00:16<00:25, 18.86it/s]
 39%|###8      | 302/781 [00:16<00:25, 18.86it/s]
 39%|###8      | 304/781 [00:16<00:25, 18.95it/s]
 39%|###9      | 306/781 [00:16<00:25, 18.96it/s]
 39%|###9      | 308/781 [00:16<00:25, 18.86it/s]
 40%|###9      | 310/781 [00:16<00:25, 18.60it/s]
 40%|###9      | 312/781 [00:16<00:25, 18.46it/s]
 40%|####      | 314/781 [00:17<00:25, 18.57it/s]
 40%|####      | 316/781 [00:17<00:24, 18.76it/s]
 41%|####      | 318/781 [00:17<00:24, 18.75it/s]
 41%|####      | 320/781 [00:17<00:24, 18.68it/s]
 41%|####1     | 322/781 [00:17<00:24, 18.52it/s]
 41%|####1     | 324/781 [00:17<00:24, 18.55it/s]
 42%|####1     | 326/781 [00:17<00:24, 18.57it/s]
 42%|####1     | 328/781 [00:17<00:24, 18.60it/s]
 42%|####2     | 330/781 [00:17<00:24, 18.68it/s]
 43%|####2     | 332/781 [00:18<00:24, 18.66it/s]
 43%|####2     | 334/781 [00:18<00:24, 18.53it/s]
 43%|####3     | 336/781 [00:18<00:23, 18.61it/s]
 43%|####3     | 338/781 [00:18<00:23, 18.62it/s]
 44%|####3     | 340/781 [00:18<00:23, 18.61it/s]
 44%|####3     | 342/781 [00:18<00:23, 18.76it/s]
 44%|####4     | 344/781 [00:18<00:23, 18.86it/s]
 44%|####4     | 346/781 [00:18<00:22, 18.92it/s]
 45%|####4     | 348/781 [00:18<00:22, 18.97it/s]
 45%|####4     | 350/781 [00:18<00:22, 18.98it/s]
 45%|####5     | 352/781 [00:19<00:22, 18.97it/s]
 45%|####5     | 354/781 [00:19<00:22, 18.93it/s]
 46%|####5     | 356/781 [00:19<00:22, 18.96it/s]
 46%|####5     | 358/781 [00:19<00:22, 18.90it/s]
 46%|####6     | 360/781 [00:19<00:22, 18.85it/s]
 46%|####6     | 362/781 [00:19<00:22, 18.89it/s]
 47%|####6     | 364/781 [00:19<00:22, 18.88it/s]
 47%|####6     | 366/781 [00:19<00:22, 18.77it/s]
 47%|####7     | 368/781 [00:19<00:21, 18.82it/s]
 47%|####7     | 370/781 [00:20<00:21, 18.73it/s]
 48%|####7     | 372/781 [00:20<00:21, 18.80it/s]
 48%|####7     | 374/781 [00:20<00:21, 18.72it/s]
 48%|####8     | 376/781 [00:20<00:21, 18.65it/s]
 48%|####8     | 378/781 [00:20<00:22, 18.18it/s]
 49%|####8     | 380/781 [00:20<00:21, 18.40it/s]
 49%|####8     | 382/781 [00:20<00:21, 18.51it/s]
 49%|####9     | 384/781 [00:20<00:21, 18.71it/s]
 49%|####9     | 386/781 [00:20<00:21, 18.80it/s]
 50%|####9     | 388/781 [00:21<00:20, 18.80it/s]
 50%|####9     | 390/781 [00:21<00:20, 18.62it/s]
 50%|#####     | 392/781 [00:21<00:20, 18.66it/s]
 50%|#####     | 394/781 [00:21<00:20, 18.70it/s]
 51%|#####     | 396/781 [00:21<00:20, 18.69it/s]
 51%|#####     | 398/781 [00:21<00:20, 18.81it/s]
 51%|#####1    | 400/781 [00:21<00:20, 18.91it/s]
 51%|#####1    | 402/781 [00:21<00:19, 18.99it/s]
 52%|#####1    | 404/781 [00:21<00:19, 18.93it/s]
 52%|#####1    | 406/781 [00:21<00:19, 18.89it/s]
 52%|#####2    | 408/781 [00:22<00:19, 18.95it/s]
 52%|#####2    | 410/781 [00:22<00:19, 18.92it/s]
 53%|#####2    | 412/781 [00:22<00:19, 18.77it/s]
 53%|#####3    | 414/781 [00:22<00:19, 18.84it/s]
 53%|#####3    | 416/781 [00:22<00:19, 18.86it/s]
 54%|#####3    | 418/781 [00:22<00:19, 18.88it/s]
 54%|#####3    | 420/781 [00:22<00:19, 18.95it/s]
 54%|#####4    | 422/781 [00:22<00:18, 19.01it/s]
 54%|#####4    | 424/781 [00:22<00:18, 18.82it/s]
 55%|#####4    | 426/781 [00:23<00:18, 18.78it/s]
 55%|#####4    | 428/781 [00:23<00:18, 18.68it/s]
 55%|#####5    | 430/781 [00:23<00:18, 18.70it/s]
 55%|#####5    | 432/781 [00:23<00:18, 18.48it/s]
 56%|#####5    | 434/781 [00:23<00:18, 18.68it/s]
 56%|#####5    | 436/781 [00:23<00:18, 18.64it/s]
 56%|#####6    | 438/781 [00:23<00:18, 18.81it/s]
 56%|#####6    | 440/781 [00:23<00:18, 18.76it/s]
 57%|#####6    | 442/781 [00:23<00:18, 18.81it/s]
 57%|#####6    | 444/781 [00:23<00:18, 18.71it/s]
 57%|#####7    | 446/781 [00:24<00:17, 18.84it/s]
 57%|#####7    | 448/781 [00:24<00:17, 18.85it/s]
 58%|#####7    | 450/781 [00:24<00:17, 18.86it/s]
 58%|#####7    | 452/781 [00:24<00:17, 18.73it/s]
 58%|#####8    | 454/781 [00:24<00:17, 18.87it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.93it/s]
 59%|#####8    | 458/781 [00:24<00:16, 19.01it/s]
 59%|#####8    | 460/781 [00:24<00:16, 19.07it/s]
 59%|#####9    | 462/781 [00:24<00:16, 18.97it/s]
 59%|#####9    | 464/781 [00:25<00:16, 18.99it/s]
 60%|#####9    | 466/781 [00:25<00:16, 19.03it/s]
 60%|#####9    | 468/781 [00:25<00:16, 18.92it/s]
 60%|######    | 470/781 [00:25<00:16, 18.93it/s]
 60%|######    | 472/781 [00:25<00:16, 18.91it/s]
 61%|######    | 474/781 [00:25<00:16, 18.87it/s]
 61%|######    | 476/781 [00:25<00:16, 18.84it/s]
 61%|######1   | 478/781 [00:25<00:15, 18.94it/s]
 61%|######1   | 480/781 [00:25<00:15, 18.88it/s]
 62%|######1   | 482/781 [00:25<00:15, 18.84it/s]
 62%|######1   | 484/781 [00:26<00:15, 18.79it/s]
 62%|######2   | 486/781 [00:26<00:15, 18.60it/s]
 62%|######2   | 488/781 [00:26<00:15, 18.54it/s]
 63%|######2   | 490/781 [00:26<00:15, 18.43it/s]
 63%|######2   | 492/781 [00:26<00:15, 18.57it/s]
 63%|######3   | 494/781 [00:26<00:15, 18.55it/s]
 64%|######3   | 496/781 [00:26<00:15, 18.68it/s]
 64%|######3   | 498/781 [00:26<00:15, 18.75it/s]
 64%|######4   | 500/781 [00:26<00:15, 18.69it/s]
 64%|######4   | 502/781 [00:27<00:14, 18.72it/s]
 65%|######4   | 504/781 [00:27<00:15, 18.35it/s]
 65%|######4   | 506/781 [00:27<00:14, 18.58it/s]
 65%|######5   | 508/781 [00:27<00:14, 18.69it/s]
 65%|######5   | 510/781 [00:27<00:14, 18.86it/s]
 66%|######5   | 512/781 [00:27<00:14, 18.77it/s]
 66%|######5   | 514/781 [00:27<00:14, 18.63it/s]
 66%|######6   | 516/781 [00:27<00:14, 18.74it/s]
 66%|######6   | 518/781 [00:27<00:13, 18.82it/s]
 67%|######6   | 520/781 [00:28<00:13, 18.89it/s]
 67%|######6   | 522/781 [00:28<00:13, 18.91it/s]
 67%|######7   | 524/781 [00:28<00:13, 18.78it/s]
 67%|######7   | 526/781 [00:28<00:13, 18.62it/s]
 68%|######7   | 528/781 [00:28<00:13, 18.56it/s]
 68%|######7   | 530/781 [00:28<00:13, 18.61it/s]
 68%|######8   | 532/781 [00:28<00:13, 18.64it/s]
 68%|######8   | 534/781 [00:28<00:13, 18.64it/s]
 69%|######8   | 536/781 [00:28<00:13, 18.67it/s]
 69%|######8   | 538/781 [00:28<00:12, 18.81it/s]
 69%|######9   | 540/781 [00:29<00:12, 18.83it/s]
 69%|######9   | 542/781 [00:29<00:12, 18.58it/s]
 70%|######9   | 544/781 [00:29<00:12, 18.64it/s]
 70%|######9   | 546/781 [00:29<00:12, 18.79it/s]
 70%|#######   | 548/781 [00:29<00:12, 18.50it/s]
 70%|#######   | 550/781 [00:29<00:12, 18.57it/s]
 71%|#######   | 552/781 [00:29<00:12, 18.75it/s]
 71%|#######   | 554/781 [00:29<00:12, 18.81it/s]
 71%|#######1  | 556/781 [00:29<00:11, 18.90it/s]
 71%|#######1  | 558/781 [00:30<00:11, 18.99it/s]
 72%|#######1  | 560/781 [00:30<00:11, 19.04it/s]
 72%|#######1  | 562/781 [00:30<00:11, 18.99it/s]
 72%|#######2  | 564/781 [00:30<00:11, 18.88it/s]
 72%|#######2  | 566/781 [00:30<00:11, 18.77it/s]
 73%|#######2  | 568/781 [00:30<00:11, 18.90it/s]
 73%|#######2  | 570/781 [00:30<00:11, 18.98it/s]
 73%|#######3  | 572/781 [00:30<00:11, 18.96it/s]
 73%|#######3  | 574/781 [00:30<00:10, 19.01it/s]
 74%|#######3  | 576/781 [00:31<00:10, 19.01it/s]
 74%|#######4  | 578/781 [00:31<00:10, 19.07it/s]
 74%|#######4  | 580/781 [00:31<00:10, 19.10it/s]
 75%|#######4  | 582/781 [00:31<00:10, 19.09it/s]
 75%|#######4  | 584/781 [00:31<00:10, 19.08it/s]
 75%|#######5  | 586/781 [00:31<00:10, 19.00it/s]
 75%|#######5  | 588/781 [00:31<00:10, 18.99it/s]
 76%|#######5  | 590/781 [00:31<00:10, 19.03it/s]
 76%|#######5  | 592/781 [00:31<00:10, 18.84it/s]
 76%|#######6  | 594/781 [00:31<00:09, 18.90it/s]
 76%|#######6  | 596/781 [00:32<00:09, 18.91it/s]
 77%|#######6  | 598/781 [00:32<00:09, 18.96it/s]
 77%|#######6  | 600/781 [00:32<00:09, 19.00it/s]
 77%|#######7  | 602/781 [00:32<00:09, 18.84it/s]
 77%|#######7  | 604/781 [00:32<00:09, 18.82it/s]
 78%|#######7  | 606/781 [00:32<00:09, 18.79it/s]
 78%|#######7  | 608/781 [00:32<00:09, 18.63it/s]
 78%|#######8  | 610/781 [00:32<00:09, 18.71it/s]
 78%|#######8  | 612/781 [00:32<00:08, 18.89it/s]
 79%|#######8  | 614/781 [00:33<00:08, 18.70it/s]
 79%|#######8  | 616/781 [00:33<00:08, 18.92it/s]
 79%|#######9  | 618/781 [00:33<00:08, 18.87it/s]
 79%|#######9  | 620/781 [00:33<00:08, 18.83it/s]
 80%|#######9  | 622/781 [00:33<00:08, 18.98it/s]
 80%|#######9  | 624/781 [00:33<00:08, 19.11it/s]
 80%|########  | 626/781 [00:33<00:08, 18.97it/s]
 80%|########  | 628/781 [00:33<00:08, 18.99it/s]
 81%|########  | 630/781 [00:33<00:07, 19.12it/s]
 81%|########  | 632/781 [00:33<00:07, 19.21it/s]
 81%|########1 | 634/781 [00:34<00:07, 19.33it/s]
 81%|########1 | 636/781 [00:34<00:07, 18.99it/s]
 82%|########1 | 638/781 [00:34<00:07, 19.15it/s]
 82%|########1 | 640/781 [00:34<00:07, 19.27it/s]
 82%|########2 | 642/781 [00:34<00:07, 19.22it/s]
 82%|########2 | 644/781 [00:34<00:07, 19.30it/s]
 83%|########2 | 646/781 [00:34<00:06, 19.34it/s]
 83%|########2 | 648/781 [00:34<00:06, 19.45it/s]
 83%|########3 | 650/781 [00:34<00:06, 19.46it/s]
 83%|########3 | 652/781 [00:34<00:06, 19.44it/s]
 84%|########3 | 654/781 [00:35<00:06, 19.46it/s]
 84%|########3 | 656/781 [00:35<00:06, 19.49it/s]
 84%|########4 | 658/781 [00:35<00:06, 19.56it/s]
 85%|########4 | 660/781 [00:35<00:06, 19.67it/s]
 85%|########4 | 662/781 [00:35<00:06, 19.60it/s]
 85%|########5 | 664/781 [00:35<00:05, 19.55it/s]
 85%|########5 | 666/781 [00:35<00:05, 19.45it/s]
 86%|########5 | 668/781 [00:35<00:05, 19.09it/s]
 86%|########5 | 670/781 [00:35<00:05, 19.07it/s]
 86%|########6 | 672/781 [00:36<00:05, 18.70it/s]
 86%|########6 | 674/781 [00:36<00:05, 18.92it/s]
 87%|########6 | 676/781 [00:36<00:05, 18.84it/s]
 87%|########6 | 678/781 [00:36<00:05, 18.82it/s]
 87%|########7 | 680/781 [00:36<00:05, 18.90it/s]
 87%|########7 | 682/781 [00:36<00:05, 18.93it/s]
 88%|########7 | 684/781 [00:36<00:05, 19.08it/s]
 88%|########7 | 686/781 [00:36<00:05, 18.83it/s]
 88%|########8 | 688/781 [00:36<00:04, 18.98it/s]
 88%|########8 | 690/781 [00:36<00:04, 18.99it/s]
 89%|########8 | 692/781 [00:37<00:04, 18.93it/s]
 89%|########8 | 694/781 [00:37<00:04, 19.07it/s]
 89%|########9 | 696/781 [00:37<00:04, 19.03it/s]
 89%|########9 | 698/781 [00:37<00:04, 18.99it/s]
 90%|########9 | 700/781 [00:37<00:04, 19.15it/s]
 90%|########9 | 702/781 [00:37<00:04, 19.01it/s]
 90%|######### | 704/781 [00:37<00:04, 19.15it/s]
 90%|######### | 706/781 [00:37<00:03, 19.07it/s]
 91%|######### | 708/781 [00:37<00:03, 19.10it/s]
 91%|######### | 710/781 [00:38<00:03, 19.11it/s]
 91%|#########1| 712/781 [00:38<00:03, 19.18it/s]
 91%|#########1| 714/781 [00:38<00:03, 19.22it/s]
 92%|#########1| 716/781 [00:38<00:03, 19.18it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.20it/s]
 92%|#########2| 720/781 [00:38<00:03, 19.29it/s]
 92%|#########2| 722/781 [00:38<00:03, 19.36it/s]
 93%|#########2| 724/781 [00:38<00:02, 19.06it/s]
 93%|#########2| 726/781 [00:38<00:02, 19.17it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.23it/s]
 93%|#########3| 730/781 [00:39<00:02, 19.35it/s]
 94%|#########3| 732/781 [00:39<00:02, 19.40it/s]
 94%|#########3| 734/781 [00:39<00:02, 19.28it/s]
 94%|#########4| 736/781 [00:39<00:02, 19.07it/s]
 94%|#########4| 738/781 [00:39<00:02, 19.04it/s]
 95%|#########4| 740/781 [00:39<00:02, 19.13it/s]
 95%|#########5| 742/781 [00:39<00:02, 19.04it/s]
 95%|#########5| 744/781 [00:39<00:01, 19.21it/s]
 96%|#########5| 746/781 [00:39<00:01, 19.32it/s]
 96%|#########5| 748/781 [00:40<00:01, 19.45it/s]
 96%|#########6| 750/781 [00:40<00:01, 19.54it/s]
 96%|#########6| 752/781 [00:40<00:01, 19.28it/s]
 97%|#########6| 754/781 [00:40<00:01, 19.08it/s]
 97%|#########6| 756/781 [00:40<00:01, 19.28it/s]
 97%|#########7| 758/781 [00:40<00:01, 19.40it/s]
 97%|#########7| 760/781 [00:40<00:01, 19.44it/s]
 98%|#########7| 762/781 [00:40<00:00, 19.52it/s]
 98%|#########7| 764/781 [00:40<00:00, 19.24it/s]
 98%|#########8| 766/781 [00:41<00:01, 13.59it/s]
 98%|#########8| 768/781 [00:41<00:00, 14.92it/s]
 99%|#########8| 770/781 [00:41<00:00, 16.06it/s]
 99%|#########8| 772/781 [00:41<00:00, 16.85it/s]
 99%|#########9| 774/781 [00:41<00:00, 17.48it/s]
 99%|#########9| 776/781 [00:41<00:00, 18.07it/s]
100%|#########9| 778/781 [00:41<00:00, 18.47it/s]
100%|#########9| 780/781 [00:41<00:00, 18.55it/s]
100%|##########| 781/781 [00:41<00:00, 18.66it/s]






=======================================================   RESTART [11-24 10:12:30]   =======================================================







=======================================================   RESTART [11-24 10:12:30]   =======================================================







=======================================================   RESTART [11-24 10:12:30]   =======================================================







=======================================================   RESTART [11-24 10:12:30]   =======================================================







=======================================================   RESTART [11-24 10:12:30]   =======================================================







=======================================================   RESTART [11-24 10:12:30]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241123_181547-szr3io5m
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run valiant-bee-543
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/szr3io5m
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(4560x28, 28x1920)
  triton_mm_11 0.0296 ms 100.0%
  triton_mm_9 0.0296 ms 99.9%
  triton_mm_6 0.0297 ms 99.7%
  triton_mm_7 0.0307 ms 96.6%
  triton_mm_8 0.0308 ms 96.2%
  triton_mm_12 0.0308 ms 96.0%
  triton_mm_4 0.0315 ms 94.0%
  triton_mm_15 0.0316 ms 93.7%
  triton_mm_3 0.0322 ms 91.9%
  triton_mm_16 0.0329 ms 90.0%
SingleProcess AUTOTUNE benchmarking takes 2.0612 seconds and 0.0067 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_29 0.0472 ms 100.0%
  triton_mm_33 0.0480 ms 98.4%
  triton_mm_25 0.0482 ms 98.1%
  triton_mm_28 0.0489 ms 96.6%
  triton_mm_24 0.0491 ms 96.3%
  triton_mm_18 0.0497 ms 95.1%
  triton_mm_21 0.0501 ms 94.3%
  bias_addmm 0.0515 ms 91.7%
  triton_mm_19 0.0517 ms 91.4%
  triton_mm_20 0.0539 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.2247 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_51 0.4410 ms 100.0%
  triton_mm_50 0.4493 ms 98.2%
  triton_mm_52 0.4883 ms 90.3%
  bias_addmm 0.5224 ms 84.4%
  triton_mm_45 0.5447 ms 81.0%
  triton_mm_47 0.5642 ms 78.2%
  triton_mm_43 0.5780 ms 76.3%
  triton_mm_46 0.5969 ms 73.9%
  triton_mm_48 0.6370 ms 69.2%
  triton_mm_44 0.6513 ms 67.7%
SingleProcess AUTOTUNE benchmarking takes 2.7246 seconds and 0.0205 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_70 0.1497 ms 100.0%
  triton_mm_69 0.1523 ms 98.3%
  triton_mm_71 0.1613 ms 92.8%
  bias_addmm 0.1651 ms 90.7%
  triton_mm_64 0.1821 ms 82.2%
  addmm 0.1884 ms 79.5%
  triton_mm_62 0.1893 ms 79.1%
  triton_mm_66 0.1947 ms 76.9%
  triton_mm_65 0.2020 ms 74.1%
  triton_mm_67 0.2120 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.5875 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(4576x7680, 4576x1920, 1920x7680)
  triton_mm_89 0.5867 ms 100.0%
  triton_mm_88 0.5965 ms 98.4%
  triton_mm_90 0.6488 ms 90.4%
  bias_addmm 0.6492 ms 90.4%
  triton_mm_83 0.7208 ms 81.4%
  triton_mm_85 0.7469 ms 78.5%
  triton_mm_81 0.7632 ms 76.9%
  addmm 0.7783 ms 75.4%
  triton_mm_84 0.7936 ms 73.9%
  triton_mm_86 0.8492 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.7583 seconds and 0.7490 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_108 0.5585 ms 100.0%
  triton_mm_107 0.5808 ms 96.2%
  triton_mm_109 0.5909 ms 94.5%
  bias_addmm 0.6200 ms 90.1%
  addmm 0.6322 ms 88.3%
  triton_mm_102 0.7107 ms 78.6%
  triton_mm_103 0.7444 ms 75.0%
  triton_mm_100 0.7484 ms 74.6%
  triton_mm_104 0.7509 ms 74.4%
  triton_mm_105 0.8064 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.7461 seconds and 0.2130 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_122 0.0473 ms 100.0%
  triton_mm_118 0.0480 ms 98.4%
  triton_mm_126 0.0488 ms 96.9%
  triton_mm_121 0.0498 ms 94.9%
  triton_mm_117 0.0500 ms 94.6%
  triton_mm_114 0.0501 ms 94.4%
  triton_mm_111 0.0501 ms 94.3%
  bias_addmm 0.0508 ms 93.0%
  triton_mm_112 0.0511 ms 92.4%
  triton_mm_113 0.0534 ms 88.5%
SingleProcess AUTOTUNE benchmarking takes 2.2187 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_201 0.5580 ms 100.0%
  triton_mm_200 0.5811 ms 96.0%
  triton_mm_202 0.5905 ms 94.5%
  bias_addmm 0.6191 ms 90.1%
  addmm 0.6678 ms 83.6%
  triton_mm_195 0.7091 ms 78.7%
  triton_mm_196 0.7449 ms 74.9%
  triton_mm_193 0.7474 ms 74.7%
  triton_mm_197 0.7528 ms 74.1%
  triton_mm_198 0.8035 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.7508 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_215 0.0474 ms 100.0%
  triton_mm_219 0.0479 ms 98.9%
  triton_mm_211 0.0482 ms 98.4%
  triton_mm_214 0.0491 ms 96.6%
  triton_mm_204 0.0498 ms 95.2%
  triton_mm_207 0.0501 ms 94.6%
  triton_mm_210 0.0501 ms 94.6%
  triton_mm_205 0.0520 ms 91.1%
  bias_addmm 0.0524 ms 90.5%
  triton_mm_206 0.0530 ms 89.5%
SingleProcess AUTOTUNE benchmarking takes 2.2230 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_294 0.5561 ms 100.0%
  triton_mm_293 0.5805 ms 95.8%
  triton_mm_295 0.5894 ms 94.4%
  bias_addmm 0.6192 ms 89.8%
  addmm 0.6317 ms 88.0%
  triton_mm_288 0.7090 ms 78.4%
  triton_mm_289 0.7438 ms 74.8%
  triton_mm_286 0.7459 ms 74.5%
  triton_mm_290 0.7500 ms 74.1%
  triton_mm_291 0.8023 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.7590 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x1920, 1920x2048)
  triton_mm_591 0.1706 ms 100.0%
  triton_mm_592 0.1706 ms 100.0%
  triton_mm_593 0.1846 ms 92.4%
  triton_mm_586 0.1912 ms 89.2%
  triton_mm_584 0.1981 ms 86.1%
  triton_mm_588 0.1997 ms 85.4%
  bias_addmm 0.2028 ms 84.1%
  triton_mm_587 0.2123 ms 80.4%
  triton_mm_589 0.2292 ms 74.4%
  triton_mm_585 0.2310 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 2.6357 seconds and 0.0161 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x2048, 2048x2048)
  triton_mm_610 0.1782 ms 100.0%
  triton_mm_611 0.1791 ms 99.5%
  bias_addmm 0.1811 ms 98.4%
  triton_mm_612 0.1933 ms 92.2%
  triton_mm_603 0.2027 ms 87.9%
  triton_mm_607 0.2030 ms 87.8%
  triton_mm_605 0.2036 ms 87.5%
  addmm 0.2087 ms 85.4%
  triton_mm_606 0.2245 ms 79.4%
  triton_mm_608 0.2392 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 11.4878 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x768)
  mm 0.0645 ms 100.0%
  triton_mm_630 0.0649 ms 99.3%
  triton_mm_629 0.0651 ms 99.2%
  triton_mm_631 0.0684 ms 94.3%
  triton_mm_625 0.0884 ms 73.0%
  triton_mm_627 0.0918 ms 70.2%
  triton_mm_623 0.0920 ms 70.2%
  triton_mm_626 0.0928 ms 69.5%
  triton_mm_622 0.0933 ms 69.2%
  triton_mm_624 0.0992 ms 65.0%
SingleProcess AUTOTUNE benchmarking takes 2.4107 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(16x3840, 16x1920, 1920x3840)
  triton_mm_2876 0.0215 ms 100.0%
  triton_mm_2872 0.0226 ms 95.0%
  triton_mm_2880 0.0227 ms 94.6%
  triton_mm_2868 0.0234 ms 91.7%
  bias_addmm 0.0251 ms 85.5%
  triton_mm_2865 0.0269 ms 79.9%
  triton_mm_2867 0.0272 ms 78.9%
  triton_mm_2871 0.0273 ms 78.6%
  triton_mm_2875 0.0274 ms 78.3%
  addmm 0.0280 ms 76.6%
SingleProcess AUTOTUNE benchmarking takes 8.6747 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x32768, 4576x1920, 1920x32768)
  triton_mm_2898 2.4926 ms 100.0%
  triton_mm_2897 2.5374 ms 98.2%
  bias_addmm 2.5993 ms 95.9%
  triton_mm_2899 2.7441 ms 90.8%
  addmm 2.9445 ms 84.7%
  triton_mm_2892 3.0507 ms 81.7%
  triton_mm_2894 3.1326 ms 79.6%
  triton_mm_2890 3.2391 ms 77.0%
  triton_mm_2893 3.3100 ms 75.3%
  triton_mm_2895 3.6256 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 12.2746 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x4576, 4576x1920)
  mm 2.3111 ms 100.0%
  triton_mm_2935 2.3132 ms 99.9%
  triton_mm_2936 2.3937 ms 96.5%
  triton_mm_2937 2.5277 ms 91.4%
  triton_mm_2928 2.9820 ms 77.5%
  triton_mm_2932 3.0197 ms 76.5%
  triton_mm_2930 3.0805 ms 75.0%
  triton_mm_2931 3.1707 ms 72.9%
  triton_mm_2933 3.2114 ms 72.0%
  triton_mm_2929 3.2565 ms 71.0%
SingleProcess AUTOTUNE benchmarking takes 9.9320 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x32768, 32768x1920)
  mm 2.2964 ms 100.0%
  triton_mm_2916 2.3671 ms 97.0%
  triton_mm_2917 2.4016 ms 95.6%
  triton_mm_2918 2.4370 ms 94.2%
  triton_mm_2912 3.0519 ms 75.2%
  triton_mm_2910 3.2627 ms 70.4%
  triton_mm_2911 3.2804 ms 70.0%
  triton_mm_2909 3.3141 ms 69.3%
  triton_mm_2914 3.3969 ms 67.6%
  triton_mm_2913 3.4950 ms 65.7%
SingleProcess AUTOTUNE benchmarking takes 12.4903 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3006 0.5576 ms 100.0%
  mm 0.5782 ms 96.4%
  triton_mm_3007 0.5820 ms 95.8%
  triton_mm_3008 0.6109 ms 91.3%
  triton_mm_3003 0.6842 ms 81.5%
  triton_mm_2999 0.7051 ms 79.1%
  triton_mm_3001 0.7197 ms 77.5%
  triton_mm_3004 0.7525 ms 74.1%
  triton_mm_3002 0.7545 ms 73.9%
  triton_mm_3000 0.7643 ms 73.0%
SingleProcess AUTOTUNE benchmarking takes 9.6593 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x16, 16x1920)
  triton_mm_2962 0.0164 ms 100.0%
  triton_mm_2964 0.0165 ms 99.2%
  triton_mm_2960 0.0175 ms 93.4%
  triton_mm_2967 0.0178 ms 91.9%
  triton_mm_2961 0.0178 ms 91.7%
  triton_mm_2963 0.0178 ms 91.7%
  triton_mm_2965 0.0179 ms 91.6%
  triton_mm_2966 0.0179 ms 91.4%
  triton_mm_2969 0.0181 ms 90.1%
  triton_mm_2957 0.0183 ms 89.5%
SingleProcess AUTOTUNE benchmarking takes 5.2284 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x3840, 3840x1920)
  triton_mm_2946 0.0229 ms 100.0%
  triton_mm_2942 0.0239 ms 95.7%
  triton_mm_2950 0.0257 ms 89.0%
  mm 0.0277 ms 82.5%
  triton_mm_2954 0.0284 ms 80.5%
  triton_mm_2940 0.0349 ms 65.5%
  triton_mm_2941 0.0350 ms 65.4%
  triton_mm_2945 0.0372 ms 61.5%
  triton_mm_2949 0.0394 ms 58.0%
  triton_mm_2939 0.0405 ms 56.5%
SingleProcess AUTOTUNE benchmarking takes 6.4494 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_2987 0.5590 ms 100.0%
  mm 0.5642 ms 99.1%
  triton_mm_2988 0.5692 ms 98.2%
  triton_mm_2989 0.6133 ms 91.1%
  triton_mm_2984 0.7071 ms 79.0%
  triton_mm_2982 0.7074 ms 79.0%
  triton_mm_2980 0.7160 ms 78.1%
  triton_mm_2983 0.7582 ms 73.7%
  triton_mm_2981 0.8002 ms 69.9%
  triton_mm_2985 0.8097 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 10.2367 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5527 ms 100.0%
  triton_mm_3044 0.5567 ms 99.3%
  triton_mm_3045 0.5841 ms 94.6%
  triton_mm_3046 0.6173 ms 89.5%
  triton_mm_3037 0.6887 ms 80.2%
  triton_mm_3041 0.6964 ms 79.4%
  triton_mm_3039 0.7113 ms 77.7%
  triton_mm_3040 0.7456 ms 74.1%
  triton_mm_3042 0.7566 ms 73.0%
  triton_mm_3038 0.7615 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 9.8093 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3025 0.5493 ms 100.0%
  triton_mm_3026 0.5686 ms 96.6%
  triton_mm_3027 0.5817 ms 94.4%
  mm 0.5890 ms 93.3%
  triton_mm_3020 0.6993 ms 78.6%
  triton_mm_3021 0.7232 ms 76.0%
  triton_mm_3018 0.7234 ms 75.9%
  triton_mm_3022 0.7328 ms 75.0%
  triton_mm_3019 0.7604 ms 72.2%
  triton_mm_3023 0.7890 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 10.1466 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4221 ms 100.0%
  triton_mm_3120 0.4293 ms 98.3%
  triton_mm_3121 0.4521 ms 93.4%
  triton_mm_3122 0.4744 ms 89.0%
  triton_mm_3117 0.5263 ms 80.2%
  triton_mm_3113 0.5272 ms 80.1%
  triton_mm_3115 0.5389 ms 78.3%
  triton_mm_3116 0.5653 ms 74.7%
  triton_mm_3118 0.5785 ms 73.0%
  triton_mm_3114 0.5791 ms 72.9%
SingleProcess AUTOTUNE benchmarking takes 10.0063 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3102 0.4135 ms 100.0%
  triton_mm_3101 0.4140 ms 99.9%
  triton_mm_3103 0.4353 ms 95.0%
  mm 0.4412 ms 93.7%
  triton_mm_3096 0.5203 ms 79.5%
  triton_mm_3094 0.5426 ms 76.2%
  triton_mm_3097 0.5455 ms 75.8%
  triton_mm_3098 0.5504 ms 75.1%
  triton_mm_3095 0.5700 ms 72.5%
  triton_mm_3099 0.5917 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 10.1146 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1644 ms 100.0%
  triton_mm_3082 0.1854 ms 88.6%
  triton_mm_3083 0.1940 ms 84.7%
  triton_mm_3077 0.1944 ms 84.5%
  triton_mm_3075 0.2058 ms 79.9%
  triton_mm_3076 0.2207 ms 74.5%
  triton_mm_3080 0.2211 ms 74.3%
  triton_mm_3084 0.2335 ms 70.4%
  triton_mm_3074 0.2347 ms 70.0%
  triton_mm_3079 0.2351 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 12.5005 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1395 ms 100.0%
  triton_mm_3063 0.1426 ms 97.8%
  triton_mm_3064 0.1433 ms 97.3%
  triton_mm_3065 0.1533 ms 90.9%
  triton_mm_3058 0.1768 ms 78.9%
  triton_mm_3056 0.1828 ms 76.3%
  triton_mm_3060 0.1849 ms 75.4%
  triton_mm_3059 0.1917 ms 72.7%
  triton_mm_3057 0.1942 ms 71.8%
  triton_mm_3061 0.2027 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 10.4675 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3149 0.0328 ms 100.0%
  triton_mm_3151 0.0328 ms 100.0%
  triton_mm_3152 0.0337 ms 97.5%
  triton_mm_3154 0.0338 ms 97.3%
  triton_mm_3145 0.0342 ms 96.0%
  triton_mm_3147 0.0342 ms 96.0%
  triton_mm_3150 0.0346 ms 95.0%
  triton_mm_3148 0.0347 ms 94.6%
  triton_mm_3155 0.0351 ms 93.5%
  triton_mm_3146 0.0353 ms 93.1%
SingleProcess AUTOTUNE benchmarking takes 5.3127 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3191 0.5628 ms 100.0%
  mm 0.5796 ms 97.1%
  triton_mm_3192 0.5902 ms 95.4%
  triton_mm_3193 0.6229 ms 90.4%
  triton_mm_3188 0.6993 ms 80.5%
  triton_mm_3184 0.7210 ms 78.1%
  triton_mm_3186 0.7353 ms 76.6%
  triton_mm_3189 0.7649 ms 73.6%
  triton_mm_3187 0.7692 ms 73.2%
  triton_mm_3185 0.7761 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.6096 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3131 0.0527 ms 100.0%
  triton_mm_3127 0.0530 ms 99.5%
  mm 0.0531 ms 99.3%
  triton_mm_3135 0.0576 ms 91.6%
  triton_mm_3139 0.0682 ms 77.4%
  triton_mm_3126 0.0845 ms 62.4%
  triton_mm_3125 0.0882 ms 59.8%
  triton_mm_3130 0.0951 ms 55.5%
  triton_mm_3134 0.1013 ms 52.0%
  triton_mm_3124 0.1039 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 6.5051 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3172 0.5661 ms 100.0%
  mm 0.5673 ms 99.8%
  triton_mm_3173 0.5712 ms 99.1%
  triton_mm_3174 0.6245 ms 90.7%
  triton_mm_3167 0.7083 ms 79.9%
  triton_mm_3165 0.7171 ms 79.0%
  triton_mm_3169 0.7189 ms 78.8%
  triton_mm_3168 0.7724 ms 73.3%
  triton_mm_3166 0.8053 ms 70.3%
  triton_mm_3170 0.8270 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.6064 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5549 ms 100.0%
  triton_mm_3229 0.5642 ms 98.3%
  triton_mm_3230 0.5956 ms 93.2%
  triton_mm_3231 0.6267 ms 88.5%
  triton_mm_3222 0.7045 ms 78.8%
  triton_mm_3226 0.7103 ms 78.1%
  triton_mm_3224 0.7244 ms 76.6%
  triton_mm_3225 0.7608 ms 72.9%
  triton_mm_3227 0.7726 ms 71.8%
  triton_mm_3223 0.7749 ms 71.6%
SingleProcess AUTOTUNE benchmarking takes 2.6140 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3210 0.5581 ms 100.0%
  triton_mm_3211 0.5748 ms 97.1%
  triton_mm_3212 0.5843 ms 95.5%
  mm 0.5959 ms 93.7%
  triton_mm_3205 0.7098 ms 78.6%
  triton_mm_3203 0.7347 ms 76.0%
  triton_mm_3206 0.7370 ms 75.7%
  triton_mm_3207 0.7447 ms 74.9%
  triton_mm_3204 0.7730 ms 72.2%
  triton_mm_3208 0.8019 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.6197 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4239 ms 100.0%
  triton_mm_3305 0.4350 ms 97.5%
  triton_mm_3306 0.4593 ms 92.3%
  triton_mm_3307 0.4829 ms 87.8%
  triton_mm_3298 0.5374 ms 78.9%
  triton_mm_3302 0.5378 ms 78.8%
  triton_mm_3300 0.5483 ms 77.3%
  triton_mm_3301 0.5741 ms 73.8%
  triton_mm_3303 0.5888 ms 72.0%
  triton_mm_3299 0.5900 ms 71.9%
SingleProcess AUTOTUNE benchmarking takes 2.5816 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3286 0.4185 ms 100.0%
  triton_mm_3287 0.4203 ms 99.6%
  triton_mm_3288 0.4406 ms 95.0%
  mm 0.4452 ms 94.0%
  triton_mm_3281 0.5308 ms 78.9%
  triton_mm_3279 0.5525 ms 75.7%
  triton_mm_3282 0.5548 ms 75.4%
  triton_mm_3283 0.5585 ms 74.9%
  triton_mm_3280 0.5778 ms 72.4%
  triton_mm_3284 0.6009 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.5891 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3337 0.0326 ms 100.0%
  triton_mm_3334 0.0330 ms 98.8%
  triton_mm_3332 0.0332 ms 98.0%
  triton_mm_3335 0.0336 ms 97.0%
  triton_mm_3339 0.0338 ms 96.4%
  triton_mm_3336 0.0343 ms 95.1%
  triton_mm_3331 0.0343 ms 95.0%
  triton_mm_3330 0.0344 ms 94.6%
  triton_mm_3333 0.0348 ms 93.7%
  triton_mm_3338 0.0351 ms 92.8%
SingleProcess AUTOTUNE benchmarking takes 2.0291 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3376 0.5633 ms 100.0%
  mm 0.5821 ms 96.8%
  triton_mm_3377 0.5913 ms 95.3%
  triton_mm_3378 0.6236 ms 90.3%
  triton_mm_3373 0.6994 ms 80.5%
  triton_mm_3369 0.7193 ms 78.3%
  triton_mm_3371 0.7319 ms 77.0%
  triton_mm_3374 0.7642 ms 73.7%
  triton_mm_3372 0.7732 ms 72.9%
  triton_mm_3370 0.7768 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.6221 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0532 ms 100.0%
  triton_mm_3312 0.0533 ms 99.8%
  triton_mm_3316 0.0533 ms 99.7%
  triton_mm_3320 0.0588 ms 90.4%
  triton_mm_3324 0.0682 ms 78.0%
  triton_mm_3311 0.0846 ms 62.8%
  triton_mm_3310 0.0884 ms 60.2%
  triton_mm_3315 0.0944 ms 56.3%
  triton_mm_3319 0.1014 ms 52.4%
  triton_mm_3309 0.1040 ms 51.1%
SingleProcess AUTOTUNE benchmarking takes 2.1607 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  mm 0.5664 ms 100.0%
  triton_mm_3357 0.5665 ms 100.0%
  triton_mm_3358 0.5729 ms 98.9%
  triton_mm_3359 0.6256 ms 90.5%
  triton_mm_3352 0.7081 ms 80.0%
  triton_mm_3354 0.7184 ms 78.8%
  triton_mm_3350 0.7203 ms 78.6%
  triton_mm_3353 0.7734 ms 73.2%
  triton_mm_3351 0.8052 ms 70.3%
  triton_mm_3355 0.8257 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6193 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4225 ms 100.0%
  triton_mm_3490 0.4349 ms 97.1%
  triton_mm_3491 0.4587 ms 92.1%
  triton_mm_3492 0.4815 ms 87.7%
  triton_mm_3487 0.5352 ms 78.9%
  triton_mm_3483 0.5362 ms 78.8%
  triton_mm_3485 0.5477 ms 77.1%
  triton_mm_3486 0.5750 ms 73.5%
  triton_mm_3484 0.5864 ms 72.0%
  triton_mm_3488 0.5867 ms 72.0%
SingleProcess AUTOTUNE benchmarking takes 2.5841 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3471 0.4183 ms 100.0%
  triton_mm_3472 0.4196 ms 99.7%
  triton_mm_3473 0.4403 ms 95.0%
  mm 0.4459 ms 93.8%
  triton_mm_3466 0.5308 ms 78.8%
  triton_mm_3464 0.5507 ms 76.0%
  triton_mm_3467 0.5548 ms 75.4%
  triton_mm_3468 0.5591 ms 74.8%
  triton_mm_3465 0.5776 ms 72.4%
  triton_mm_3469 0.6024 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.5867 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0530 ms 100.0%
  triton_mm_3501 0.0531 ms 99.8%
  triton_mm_3497 0.0532 ms 99.8%
  triton_mm_3505 0.0577 ms 91.9%
  triton_mm_3509 0.0692 ms 76.6%
  triton_mm_3496 0.0838 ms 63.3%
  triton_mm_3495 0.0875 ms 60.6%
  triton_mm_3500 0.0952 ms 55.7%
  triton_mm_3504 0.1007 ms 52.7%
  triton_mm_3494 0.1032 ms 51.4%
SingleProcess AUTOTUNE benchmarking takes 2.1608 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  mm 0.5649 ms 100.0%
  triton_mm_3542 0.5667 ms 99.7%
  triton_mm_3543 0.5716 ms 98.8%
  triton_mm_3544 0.6239 ms 90.5%
  triton_mm_3537 0.7066 ms 79.9%
  triton_mm_3539 0.7139 ms 79.1%
  triton_mm_3535 0.7207 ms 78.4%
  triton_mm_3538 0.7722 ms 73.2%
  triton_mm_3536 0.8047 ms 70.2%
  triton_mm_3540 0.8267 ms 68.3%
SingleProcess AUTOTUNE benchmarking takes 2.6151 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4215 ms 100.0%
  triton_mm_3675 0.4339 ms 97.2%
  triton_mm_3676 0.4578 ms 92.1%
  triton_mm_3677 0.4819 ms 87.5%
  triton_mm_3668 0.5359 ms 78.7%
  triton_mm_3672 0.5360 ms 78.6%
  triton_mm_3670 0.5480 ms 76.9%
  triton_mm_3671 0.5724 ms 73.6%
  triton_mm_3669 0.5860 ms 71.9%
  triton_mm_3673 0.5865 ms 71.9%
SingleProcess AUTOTUNE benchmarking takes 2.5857 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3656 0.4185 ms 100.0%
  triton_mm_3657 0.4197 ms 99.7%
  triton_mm_3658 0.4409 ms 94.9%
  mm 0.4446 ms 94.1%
  triton_mm_3651 0.5299 ms 79.0%
  triton_mm_3649 0.5504 ms 76.0%
  triton_mm_3652 0.5537 ms 75.6%
  triton_mm_3653 0.5582 ms 75.0%
  triton_mm_3650 0.5764 ms 72.6%
  triton_mm_3654 0.5998 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 2.5868 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0531 ms 100.0%
  triton_mm_3686 0.0532 ms 99.8%
  triton_mm_3682 0.0536 ms 99.1%
  triton_mm_3690 0.0577 ms 92.0%
  triton_mm_3694 0.0693 ms 76.5%
  triton_mm_3681 0.0838 ms 63.3%
  triton_mm_3680 0.0875 ms 60.6%
  triton_mm_3685 0.0953 ms 55.7%
  triton_mm_3689 0.1006 ms 52.7%
  triton_mm_3679 0.1032 ms 51.4%
SingleProcess AUTOTUNE benchmarking takes 2.1637 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3727 0.5640 ms 100.0%
  mm 0.5647 ms 99.9%
  triton_mm_3728 0.5706 ms 98.8%
  triton_mm_3729 0.6226 ms 90.6%
  triton_mm_3722 0.7020 ms 80.3%
  triton_mm_3724 0.7138 ms 79.0%
  triton_mm_3720 0.7169 ms 78.7%
  triton_mm_3723 0.7712 ms 73.1%
  triton_mm_3721 0.8050 ms 70.1%
  triton_mm_3725 0.8250 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.6187 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4254 ms 100.0%
  triton_mm_3860 0.4329 ms 98.3%
  triton_mm_3861 0.4558 ms 93.3%
  triton_mm_3862 0.4778 ms 89.0%
  triton_mm_3853 0.5325 ms 79.9%
  triton_mm_3857 0.5330 ms 79.8%
  triton_mm_3855 0.5435 ms 78.3%
  triton_mm_3856 0.5712 ms 74.5%
  triton_mm_3858 0.5811 ms 73.2%
  triton_mm_3854 0.5840 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 2.5873 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3841 0.4159 ms 100.0%
  triton_mm_3842 0.4174 ms 99.6%
  triton_mm_3843 0.4376 ms 95.0%
  mm 0.4423 ms 94.0%
  triton_mm_3836 0.5266 ms 79.0%
  triton_mm_3834 0.5464 ms 76.1%
  triton_mm_3837 0.5506 ms 75.5%
  triton_mm_3838 0.5550 ms 74.9%
  triton_mm_3835 0.5737 ms 72.5%
  triton_mm_3839 0.5953 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.5866 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3867 0.0528 ms 100.0%
  triton_mm_3871 0.0530 ms 99.5%
  mm 0.0531 ms 99.4%
  triton_mm_3875 0.0575 ms 91.7%
  triton_mm_3879 0.0682 ms 77.4%
  triton_mm_3866 0.0837 ms 63.0%
  triton_mm_3865 0.0876 ms 60.3%
  triton_mm_3870 0.0944 ms 55.9%
  triton_mm_3874 0.1015 ms 52.0%
  triton_mm_3864 0.1031 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 2.1543 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3912 0.5609 ms 100.0%
  mm 0.5647 ms 99.3%
  triton_mm_3913 0.5678 ms 98.8%
  triton_mm_3914 0.6184 ms 90.7%
  triton_mm_3907 0.7026 ms 79.8%
  triton_mm_3905 0.7117 ms 78.8%
  triton_mm_3909 0.7127 ms 78.7%
  triton_mm_3908 0.7650 ms 73.3%
  triton_mm_3906 0.8050 ms 69.7%
  triton_mm_3910 0.8169 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.6188 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4246 ms 100.0%
  triton_mm_4045 0.4318 ms 98.3%
  triton_mm_4046 0.4563 ms 93.1%
  triton_mm_4047 0.4789 ms 88.7%
  triton_mm_4038 0.5311 ms 79.9%
  triton_mm_4042 0.5321 ms 79.8%
  triton_mm_4040 0.5431 ms 78.2%
  triton_mm_4041 0.5701 ms 74.5%
  triton_mm_4043 0.5829 ms 72.8%
  triton_mm_4039 0.5831 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 2.5862 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_4026 0.4160 ms 100.0%
  triton_mm_4027 0.4168 ms 99.8%
  triton_mm_4028 0.4371 ms 95.2%
  mm 0.4435 ms 93.8%
  triton_mm_4021 0.5257 ms 79.1%
  triton_mm_4019 0.5462 ms 76.2%
  triton_mm_4022 0.5511 ms 75.5%
  triton_mm_4023 0.5545 ms 75.0%
  triton_mm_4020 0.5739 ms 72.5%
  triton_mm_4024 0.5951 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.5862 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_4052 0.0529 ms 100.0%
  mm 0.0530 ms 99.8%
  triton_mm_4056 0.0534 ms 99.1%
  triton_mm_4060 0.0578 ms 91.6%
  triton_mm_4064 0.0692 ms 76.5%
  triton_mm_4051 0.0845 ms 62.6%
  triton_mm_4050 0.0882 ms 60.0%
  triton_mm_4055 0.0952 ms 55.6%
  triton_mm_4059 0.1014 ms 52.2%
  triton_mm_4049 0.1039 ms 50.9%
SingleProcess AUTOTUNE benchmarking takes 2.1744 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_4097 0.5615 ms 100.0%
  triton_mm_4098 0.5669 ms 99.0%
  mm 0.5685 ms 98.8%
  triton_mm_4099 0.6188 ms 90.7%
  triton_mm_4092 0.7014 ms 80.1%
  triton_mm_4090 0.7115 ms 78.9%
  triton_mm_4094 0.7115 ms 78.9%
  triton_mm_4093 0.7655 ms 73.4%
  triton_mm_4091 0.7972 ms 70.4%
  triton_mm_4095 0.8183 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6209 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4234 ms 100.0%
  triton_mm_4230 0.4307 ms 98.3%
  triton_mm_4231 0.4549 ms 93.1%
  triton_mm_4232 0.4777 ms 88.6%
  triton_mm_4223 0.5295 ms 80.0%
  triton_mm_4227 0.5307 ms 79.8%
  triton_mm_4225 0.5433 ms 77.9%
  triton_mm_4226 0.5703 ms 74.2%
  triton_mm_4228 0.5820 ms 72.8%
  triton_mm_4224 0.5833 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 2.5955 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_4212 0.4158 ms 100.0%
  triton_mm_4211 0.4159 ms 100.0%
  triton_mm_4213 0.4385 ms 94.8%
  mm 0.4421 ms 94.0%
  triton_mm_4206 0.5248 ms 79.2%
  triton_mm_4204 0.5465 ms 76.1%
  triton_mm_4207 0.5483 ms 75.8%
  triton_mm_4208 0.5531 ms 75.2%
  triton_mm_4205 0.5719 ms 72.7%
  triton_mm_4209 0.5938 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 2.5944 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0531 ms 100.0%
  triton_mm_4241 0.0532 ms 99.7%
  triton_mm_4237 0.0535 ms 99.2%
  triton_mm_4245 0.0587 ms 90.5%
  triton_mm_4249 0.0682 ms 77.9%
  triton_mm_4236 0.0838 ms 63.4%
  triton_mm_4235 0.0875 ms 60.7%
  triton_mm_4240 0.0945 ms 56.2%
  triton_mm_4244 0.1015 ms 52.3%
  triton_mm_4234 0.1031 ms 51.5%
SingleProcess AUTOTUNE benchmarking takes 2.1849 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_4282 0.5611 ms 100.0%
  triton_mm_4283 0.5672 ms 98.9%
  mm 0.5689 ms 98.6%
  triton_mm_4284 0.6191 ms 90.6%
  triton_mm_4277 0.6984 ms 80.3%
  triton_mm_4279 0.7111 ms 78.9%
  triton_mm_4275 0.7142 ms 78.6%
  triton_mm_4278 0.7646 ms 73.4%
  triton_mm_4276 0.8035 ms 69.8%
  triton_mm_4280 0.8184 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6286 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4241 ms 100.0%
  triton_mm_4415 0.4301 ms 98.6%
  triton_mm_4416 0.4545 ms 93.3%
  triton_mm_4417 0.4776 ms 88.8%
  triton_mm_4412 0.5304 ms 79.9%
  triton_mm_4408 0.5316 ms 79.8%
  triton_mm_4410 0.5429 ms 78.1%
  triton_mm_4411 0.5723 ms 74.1%
  triton_mm_4409 0.5813 ms 73.0%
  triton_mm_4413 0.5822 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 2.5914 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_4396 0.4159 ms 100.0%
  triton_mm_4397 0.4159 ms 100.0%
  triton_mm_4398 0.4382 ms 94.9%
  mm 0.4418 ms 94.1%
  triton_mm_4391 0.5250 ms 79.2%
  triton_mm_4389 0.5466 ms 76.1%
  triton_mm_4392 0.5486 ms 75.8%
  triton_mm_4393 0.5527 ms 75.2%
  triton_mm_4390 0.5723 ms 72.7%
  triton_mm_4394 0.5937 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 2.5985 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0531 ms 100.0%
  triton_mm_4426 0.0531 ms 100.0%
  triton_mm_4422 0.0532 ms 99.8%
  triton_mm_4430 0.0575 ms 92.4%
  triton_mm_4434 0.0695 ms 76.4%
  triton_mm_4421 0.0846 ms 62.8%
  triton_mm_4420 0.0884 ms 60.1%
  triton_mm_4425 0.0953 ms 55.7%
  triton_mm_4429 0.1014 ms 52.4%
  triton_mm_4419 0.1041 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 2.1776 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_4467 0.5606 ms 100.0%
  triton_mm_4468 0.5676 ms 98.8%
  mm 0.5685 ms 98.6%
  triton_mm_4469 0.6231 ms 90.0%
  triton_mm_4462 0.7002 ms 80.1%
  triton_mm_4464 0.7122 ms 78.7%
  triton_mm_4460 0.7134 ms 78.6%
  triton_mm_4463 0.7696 ms 72.8%
  triton_mm_4461 0.7979 ms 70.3%
  triton_mm_4465 0.8171 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6191 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_4611 0.0519 ms 100.0%
  mm 0.0531 ms 97.6%
  triton_mm_4607 0.0536 ms 96.7%
  triton_mm_4615 0.0587 ms 88.3%
  triton_mm_4619 0.0682 ms 76.0%
  triton_mm_4606 0.0847 ms 61.3%
  triton_mm_4605 0.0876 ms 59.2%
  triton_mm_4610 0.0953 ms 54.4%
  triton_mm_4614 0.1006 ms 51.5%
  triton_mm_4604 0.1032 ms 50.3%
SingleProcess AUTOTUNE benchmarking takes 2.1755 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_4652 0.5613 ms 100.0%
  triton_mm_4653 0.5665 ms 99.1%
  mm 0.5690 ms 98.6%
  triton_mm_4654 0.6186 ms 90.7%
  triton_mm_4647 0.7016 ms 80.0%
  triton_mm_4649 0.7123 ms 78.8%
  triton_mm_4645 0.7139 ms 78.6%
  triton_mm_4648 0.7648 ms 73.4%
  triton_mm_4646 0.7971 ms 70.4%
  triton_mm_4650 0.8184 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6189 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_4837 0.5615 ms 100.0%
  mm 0.5678 ms 98.9%
  triton_mm_4838 0.5687 ms 98.7%
  triton_mm_4839 0.6193 ms 90.7%
  triton_mm_4832 0.7042 ms 79.7%
  triton_mm_4834 0.7122 ms 78.8%
  triton_mm_4830 0.7158 ms 78.4%
  triton_mm_4833 0.7710 ms 72.8%
  triton_mm_4831 0.8066 ms 69.6%
  triton_mm_4835 0.8177 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.6225 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0531 ms 100.0%
  triton_mm_4977 0.0532 ms 99.9%
  triton_mm_4981 0.0534 ms 99.5%
  triton_mm_4985 0.0577 ms 92.0%
  triton_mm_4989 0.0693 ms 76.6%
  triton_mm_4976 0.0838 ms 63.3%
  triton_mm_4975 0.0875 ms 60.7%
  triton_mm_4980 0.0952 ms 55.8%
  triton_mm_4984 0.1006 ms 52.8%
  triton_mm_4974 0.1032 ms 51.4%
SingleProcess AUTOTUNE benchmarking takes 2.1639 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_5022 0.5617 ms 100.0%
  triton_mm_5023 0.5675 ms 99.0%
  mm 0.5683 ms 98.9%
  triton_mm_5024 0.6207 ms 90.5%
  triton_mm_5017 0.7014 ms 80.1%
  triton_mm_5015 0.7116 ms 78.9%
  triton_mm_5019 0.7120 ms 78.9%
  triton_mm_5018 0.7663 ms 73.3%
  triton_mm_5016 0.8051 ms 69.8%
  triton_mm_5020 0.8187 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.6167 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_5162 0.0528 ms 100.0%
  triton_mm_5166 0.0529 ms 99.8%
  mm 0.0531 ms 99.6%
  triton_mm_5170 0.0587 ms 90.0%
  triton_mm_5174 0.0682 ms 77.5%
  triton_mm_5161 0.0838 ms 63.1%
  triton_mm_5160 0.0875 ms 60.4%
  triton_mm_5165 0.0945 ms 55.9%
  triton_mm_5169 0.1007 ms 52.5%
  triton_mm_5159 0.1031 ms 51.3%
SingleProcess AUTOTUNE benchmarking takes 2.1666 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_5207 0.5625 ms 100.0%
  triton_mm_5208 0.5677 ms 99.1%
  mm 0.5694 ms 98.8%
  triton_mm_5209 0.6194 ms 90.8%
  triton_mm_5202 0.7015 ms 80.2%
  triton_mm_5204 0.7111 ms 79.1%
  triton_mm_5200 0.7120 ms 79.0%
  triton_mm_5203 0.7646 ms 73.6%
  triton_mm_5201 0.8050 ms 69.9%
  triton_mm_5205 0.8188 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.6190 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x768, 768x2048)
  triton_mm_7427 0.0752 ms 100.0%
  triton_mm_7428 0.0775 ms 97.0%
  triton_mm_7422 0.0830 ms 90.6%
  triton_mm_7420 0.0832 ms 90.4%
  triton_mm_7424 0.0833 ms 90.4%
  triton_mm_7421 0.0903 ms 83.4%
  mm 0.0924 ms 81.4%
  triton_mm_7425 0.0935 ms 80.5%
  triton_mm_7429 0.0955 ms 78.8%
  triton_mm_7423 0.1088 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.4069 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x2048)
  mm 0.1726 ms 100.0%
  triton_mm_7465 0.1735 ms 99.5%
  triton_mm_7466 0.1823 ms 94.7%
  triton_mm_7467 0.1893 ms 91.1%
  triton_mm_7462 0.2012 ms 85.8%
  triton_mm_7458 0.2028 ms 85.1%
  triton_mm_7460 0.2036 ms 84.8%
  triton_mm_7461 0.2192 ms 78.7%
  triton_mm_7459 0.2264 ms 76.2%
  triton_mm_7463 0.2320 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 10.3022 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x2048)
  mm 0.1676 ms 100.0%
  triton_mm_7484 0.1883 ms 89.0%
  triton_mm_7485 0.1947 ms 86.0%
  triton_mm_7486 0.2008 ms 83.4%
  triton_mm_7481 0.2112 ms 79.3%
  triton_mm_7477 0.2117 ms 79.1%
  triton_mm_7480 0.2187 ms 76.6%
  triton_mm_7479 0.2195 ms 76.3%
  triton_mm_7482 0.2228 ms 75.2%
  triton_mm_7478 0.2250 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 9.6575 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x1920)
  mm 0.1665 ms 100.0%
  triton_mm_7522 0.1886 ms 88.3%
  triton_mm_7523 0.1949 ms 85.4%
  triton_mm_7517 0.2012 ms 82.8%
  triton_mm_7515 0.2071 ms 80.4%
  triton_mm_7519 0.2073 ms 80.3%
  triton_mm_7518 0.2140 ms 77.8%
  triton_mm_7516 0.2231 ms 74.6%
  triton_mm_7520 0.2236 ms 74.5%
  triton_mm_7524 0.2353 ms 70.8%
SingleProcess AUTOTUNE benchmarking takes 9.7061 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x4576, 4576x2048)
  mm 0.0686 ms 100.0%
  triton_mm_7448 0.0712 ms 96.4%
  triton_mm_7447 0.0779 ms 88.2%
  triton_mm_7441 0.0814 ms 84.3%
  triton_mm_7443 0.0861 ms 79.7%
  triton_mm_7446 0.0874 ms 78.5%
  triton_mm_7439 0.0878 ms 78.2%
  triton_mm_7442 0.0901 ms 76.2%
  triton_mm_7444 0.0908 ms 75.6%
  triton_mm_7440 0.0917 ms 74.9%
SingleProcess AUTOTUNE benchmarking takes 9.6976 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x1920)
  mm 0.1484 ms 100.0%
  triton_mm_7503 0.1520 ms 97.6%
  triton_mm_7504 0.1576 ms 94.2%
  triton_mm_7505 0.1625 ms 91.3%
  triton_mm_7498 0.1898 ms 78.2%
  triton_mm_7496 0.1954 ms 76.0%
  triton_mm_7500 0.1956 ms 75.9%
  triton_mm_7499 0.2042 ms 72.7%
  triton_mm_7497 0.2070 ms 71.7%
  triton_mm_7501 0.2130 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 10.2517 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4560, 4560x28)
  triton_mm_8646 0.0330 ms 100.0%
  triton_mm_8639 0.0347 ms 95.1%
  mm 0.0363 ms 91.0%
  triton_mm_8637 0.0421 ms 78.4%
  triton_mm_8638 0.0473 ms 69.7%
  triton_mm_8645 0.0500 ms 66.0%
  triton_mm_8636 0.0507 ms 65.0%
  triton_mm_8651 0.0509 ms 64.8%
  triton_mm_8642 0.0514 ms 64.2%
  triton_mm_8650 0.0557 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 6.8903 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(4560x28, 28x1920)
  triton_mm_7 0.0296 ms 100.0%
  triton_mm_11 0.0297 ms 99.6%
  triton_mm_6 0.0298 ms 99.2%
  triton_mm_9 0.0308 ms 96.0%
  triton_mm_12 0.0309 ms 95.8%
  triton_mm_8 0.0309 ms 95.6%
  triton_mm_3 0.0320 ms 92.4%
  triton_mm_15 0.0323 ms 91.7%
  triton_mm_4 0.0326 ms 90.8%
  triton_mm_2 0.0326 ms 90.8%
SingleProcess AUTOTUNE benchmarking takes 2.0700 seconds and 0.0141 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_29 0.0474 ms 100.0%
  triton_mm_25 0.0482 ms 98.3%
  triton_mm_33 0.0487 ms 97.2%
  triton_mm_18 0.0496 ms 95.5%
  triton_mm_24 0.0501 ms 94.4%
  triton_mm_28 0.0503 ms 94.1%
  triton_mm_21 0.0503 ms 94.1%
  bias_addmm 0.0517 ms 91.6%
  triton_mm_19 0.0523 ms 90.6%
  triton_mm_20 0.0536 ms 88.4%
SingleProcess AUTOTUNE benchmarking takes 2.2432 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_51 0.4371 ms 100.0%
  triton_mm_50 0.4437 ms 98.5%
  triton_mm_52 0.4823 ms 90.6%
  bias_addmm 0.5210 ms 83.9%
  triton_mm_45 0.5360 ms 81.6%
  triton_mm_47 0.5540 ms 78.9%
  triton_mm_43 0.5732 ms 76.3%
  addmm 0.5920 ms 73.8%
  triton_mm_46 0.5962 ms 73.3%
  triton_mm_48 0.6323 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.7547 seconds and 0.0141 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_70 0.1465 ms 100.0%
  triton_mm_69 0.1510 ms 97.1%
  triton_mm_71 0.1589 ms 92.2%
  bias_addmm 0.1635 ms 89.6%
  triton_mm_64 0.1791 ms 81.8%
  addmm 0.1864 ms 78.6%
  triton_mm_62 0.1868 ms 78.4%
  triton_mm_66 0.1872 ms 78.3%
  triton_mm_65 0.1999 ms 73.3%
  triton_mm_67 0.2074 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.5838 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(4576x7680, 4576x1920, 1920x7680)
  triton_mm_89 0.5865 ms 100.0%
  triton_mm_88 0.5945 ms 98.7%
  triton_mm_90 0.6422 ms 91.3%
  bias_addmm 0.6492 ms 90.3%
  addmm 0.6780 ms 86.5%
  triton_mm_83 0.7214 ms 81.3%
  triton_mm_85 0.7381 ms 79.5%
  triton_mm_81 0.7695 ms 76.2%
  triton_mm_84 0.7935 ms 73.9%
  triton_mm_86 0.8466 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.7577 seconds and 0.2952 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_108 0.5528 ms 100.0%
  triton_mm_107 0.5755 ms 96.1%
  triton_mm_109 0.5845 ms 94.6%
  bias_addmm 0.6197 ms 89.2%
  triton_mm_102 0.7098 ms 77.9%
  addmm 0.7280 ms 75.9%
  triton_mm_103 0.7374 ms 75.0%
  triton_mm_100 0.7427 ms 74.4%
  triton_mm_104 0.7599 ms 72.8%
  triton_mm_105 0.8034 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.7521 seconds and 0.1853 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_201 0.5530 ms 100.0%
  triton_mm_200 0.5754 ms 96.1%
  triton_mm_202 0.5848 ms 94.6%
  bias_addmm 0.6203 ms 89.1%
  addmm 0.6335 ms 87.3%
  triton_mm_195 0.7088 ms 78.0%
  triton_mm_193 0.7494 ms 73.8%
  triton_mm_197 0.7500 ms 73.7%
  triton_mm_196 0.7542 ms 73.3%
  triton_mm_198 0.8030 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7571 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_294 0.5553 ms 100.0%
  triton_mm_293 0.5751 ms 96.5%
  triton_mm_295 0.5864 ms 94.7%
  bias_addmm 0.6212 ms 89.4%
  addmm 0.6340 ms 87.6%
  triton_mm_288 0.7106 ms 78.1%
  triton_mm_289 0.7372 ms 75.3%
  triton_mm_286 0.7430 ms 74.7%
  triton_mm_290 0.7501 ms 74.0%
  triton_mm_291 0.7974 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.7369 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_387 0.5557 ms 100.0%
  triton_mm_386 0.5756 ms 96.6%
  triton_mm_388 0.5879 ms 94.5%
  bias_addmm 0.6199 ms 89.6%
  addmm 0.6659 ms 83.5%
  triton_mm_381 0.7105 ms 78.2%
  triton_mm_382 0.7392 ms 75.2%
  triton_mm_379 0.7445 ms 74.6%
  triton_mm_383 0.7455 ms 74.5%
  triton_mm_384 0.8062 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7518 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_480 0.5551 ms 100.0%
  triton_mm_479 0.5755 ms 96.5%
  triton_mm_481 0.5876 ms 94.5%
  bias_addmm 0.6200 ms 89.5%
  triton_mm_474 0.7095 ms 78.2%
  addmm 0.7285 ms 76.2%
  triton_mm_475 0.7392 ms 75.1%
  triton_mm_472 0.7425 ms 74.8%
  triton_mm_476 0.7502 ms 74.0%
  triton_mm_477 0.7944 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.7535 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_573 0.5550 ms 100.0%
  triton_mm_572 0.5709 ms 97.2%
  triton_mm_574 0.5867 ms 94.6%
  bias_addmm 0.6196 ms 89.6%
  addmm 0.6269 ms 88.5%
  triton_mm_567 0.7098 ms 78.2%
  triton_mm_568 0.7367 ms 75.3%
  triton_mm_565 0.7438 ms 74.6%
  triton_mm_569 0.7508 ms 73.9%
  triton_mm_570 0.7940 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.7500 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x1920, 1920x2048)
  triton_mm_592 0.1677 ms 100.0%
  triton_mm_591 0.1697 ms 98.8%
  triton_mm_593 0.1836 ms 91.3%
  triton_mm_586 0.1887 ms 88.9%
  triton_mm_588 0.1975 ms 84.9%
  triton_mm_584 0.1980 ms 84.7%
  bias_addmm 0.2038 ms 82.3%
  triton_mm_587 0.2112 ms 79.4%
  triton_mm_585 0.2271 ms 73.9%
  triton_mm_589 0.2273 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 2.5711 seconds and 0.0122 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x2048, 2048x2048)
  triton_mm_611 0.1768 ms 100.0%
  triton_mm_610 0.1777 ms 99.5%
  bias_addmm 0.1811 ms 97.6%
  triton_mm_612 0.1926 ms 91.8%
  triton_mm_605 0.2032 ms 87.0%
  triton_mm_603 0.2034 ms 86.9%
  triton_mm_607 0.2038 ms 86.7%
  addmm 0.2083 ms 84.8%
  triton_mm_606 0.2232 ms 79.2%
  triton_mm_608 0.2382 ms 74.2%
SingleProcess AUTOTUNE benchmarking takes 2.6046 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x768)
  mm 0.0633 ms 100.0%
  triton_mm_630 0.0640 ms 98.9%
  triton_mm_629 0.0648 ms 97.7%
  triton_mm_631 0.0693 ms 91.4%
  triton_mm_625 0.0904 ms 70.0%
  triton_mm_623 0.0908 ms 69.8%
  triton_mm_627 0.0913 ms 69.4%
  triton_mm_626 0.0928 ms 68.3%
  triton_mm_622 0.0940 ms 67.3%
  triton_mm_624 0.0992 ms 63.8%
SingleProcess AUTOTUNE benchmarking takes 2.3991 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_723 0.5553 ms 100.0%
  triton_mm_722 0.5751 ms 96.5%
  triton_mm_724 0.5866 ms 94.7%
  bias_addmm 0.6204 ms 89.5%
  triton_mm_717 0.7026 ms 79.0%
  addmm 0.7284 ms 76.2%
  triton_mm_718 0.7387 ms 75.2%
  triton_mm_715 0.7441 ms 74.6%
  triton_mm_719 0.7473 ms 74.3%
  triton_mm_720 0.8060 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7611 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_816 0.5517 ms 100.0%
  triton_mm_815 0.5744 ms 96.1%
  triton_mm_817 0.5836 ms 94.5%
  bias_addmm 0.6201 ms 89.0%
  addmm 0.6853 ms 80.5%
  triton_mm_810 0.7104 ms 77.7%
  triton_mm_811 0.7364 ms 74.9%
  triton_mm_808 0.7436 ms 74.2%
  triton_mm_812 0.7453 ms 74.0%
  triton_mm_813 0.7934 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.7662 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_909 0.5558 ms 100.0%
  triton_mm_908 0.5757 ms 96.5%
  triton_mm_910 0.5872 ms 94.7%
  bias_addmm 0.6201 ms 89.6%
  addmm 0.6333 ms 87.8%
  triton_mm_903 0.7045 ms 78.9%
  triton_mm_904 0.7408 ms 75.0%
  triton_mm_901 0.7429 ms 74.8%
  triton_mm_905 0.7450 ms 74.6%
  triton_mm_906 0.8065 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.7544 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x3840, 16x1920, 1920x3840)
  triton_mm_2876 0.0214 ms 100.0%
  triton_mm_2872 0.0226 ms 94.6%
  triton_mm_2868 0.0235 ms 91.0%
  triton_mm_2880 0.0239 ms 89.3%
  bias_addmm 0.0250 ms 85.5%
  triton_mm_2865 0.0268 ms 79.6%
  triton_mm_2871 0.0272 ms 78.7%
  triton_mm_2866 0.0277 ms 77.1%
  triton_mm_2875 0.0285 ms 75.1%
  triton_mm_2867 0.0285 ms 75.0%
SingleProcess AUTOTUNE benchmarking takes 7.9611 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x32768, 4576x1920, 1920x32768)
  triton_mm_2898 2.5195 ms 100.0%
  triton_mm_2897 2.5386 ms 99.2%
  bias_addmm 2.6027 ms 96.8%
  triton_mm_2899 2.7755 ms 90.8%
  triton_mm_2892 3.0559 ms 82.4%
  triton_mm_2894 3.1631 ms 79.7%
  addmm 3.2189 ms 78.3%
  triton_mm_2890 3.2421 ms 77.7%
  triton_mm_2893 3.3124 ms 76.1%
  triton_mm_2895 3.6118 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 12.2267 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x4576, 4576x1920)
  triton_mm_2935 2.3492 ms 100.0%
  mm 2.3644 ms 99.4%
  triton_mm_2936 2.4221 ms 97.0%
  triton_mm_2937 2.5871 ms 90.8%
  triton_mm_2928 3.0609 ms 76.8%
  triton_mm_2932 3.1082 ms 75.6%
  triton_mm_2930 3.1268 ms 75.1%
  triton_mm_2931 3.2186 ms 73.0%
  triton_mm_2933 3.2995 ms 71.2%
  triton_mm_2929 3.3345 ms 70.5%
SingleProcess AUTOTUNE benchmarking takes 10.3699 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x32768, 32768x1920)
  mm 2.3118 ms 100.0%
  triton_mm_2916 2.3751 ms 97.3%
  triton_mm_2917 2.4292 ms 95.2%
  triton_mm_2918 2.4734 ms 93.5%
  triton_mm_2912 3.0988 ms 74.6%
  triton_mm_2910 3.2979 ms 70.1%
  triton_mm_2911 3.3018 ms 70.0%
  triton_mm_2909 3.3050 ms 69.9%
  triton_mm_2914 3.4330 ms 67.3%
  triton_mm_2913 3.4891 ms 66.3%
SingleProcess AUTOTUNE benchmarking takes 12.0186 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3006 0.5569 ms 100.0%
  mm 0.5777 ms 96.4%
  triton_mm_3007 0.5804 ms 96.0%
  triton_mm_3008 0.6106 ms 91.2%
  triton_mm_3003 0.6884 ms 80.9%
  triton_mm_2999 0.7082 ms 78.6%
  triton_mm_3001 0.7316 ms 76.1%
  triton_mm_3004 0.7547 ms 73.8%
  triton_mm_3002 0.7654 ms 72.8%
  triton_mm_3000 0.7753 ms 71.8%
SingleProcess AUTOTUNE benchmarking takes 10.0155 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x16, 16x1920)
  triton_mm_2960 0.0163 ms 100.0%
  triton_mm_2962 0.0163 ms 100.0%
  triton_mm_2964 0.0164 ms 99.6%
  triton_mm_2967 0.0165 ms 98.8%
  triton_mm_2961 0.0165 ms 98.5%
  triton_mm_2963 0.0166 ms 97.9%
  triton_mm_2965 0.0167 ms 97.7%
  triton_mm_2966 0.0178 ms 91.3%
  triton_mm_2959 0.0181 ms 90.1%
  triton_mm_2970 0.0185 ms 87.9%
SingleProcess AUTOTUNE benchmarking takes 5.2971 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x3840, 3840x1920)
  triton_mm_2946 0.0214 ms 100.0%
  triton_mm_2942 0.0235 ms 91.2%
  triton_mm_2950 0.0245 ms 87.5%
  mm 0.0280 ms 76.5%
  triton_mm_2954 0.0297 ms 72.2%
  triton_mm_2941 0.0342 ms 62.7%
  triton_mm_2940 0.0350 ms 61.2%
  triton_mm_2945 0.0364 ms 58.9%
  triton_mm_2949 0.0386 ms 55.6%
  triton_mm_2939 0.0413 ms 51.9%
SingleProcess AUTOTUNE benchmarking takes 6.5094 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_2987 0.5583 ms 100.0%
  triton_mm_2988 0.5636 ms 99.1%
  mm 0.5674 ms 98.4%
  triton_mm_2989 0.6136 ms 91.0%
  triton_mm_2982 0.7059 ms 79.1%
  triton_mm_2984 0.7145 ms 78.1%
  triton_mm_2980 0.7153 ms 78.1%
  triton_mm_2983 0.7635 ms 73.1%
  triton_mm_2981 0.7924 ms 70.5%
  triton_mm_2985 0.8125 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 10.3129 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5523 ms 100.0%
  triton_mm_3044 0.5563 ms 99.3%
  triton_mm_3045 0.5844 ms 94.5%
  triton_mm_3046 0.6170 ms 89.5%
  triton_mm_3037 0.6913 ms 79.9%
  triton_mm_3041 0.7037 ms 78.5%
  triton_mm_3039 0.7192 ms 76.8%
  triton_mm_3040 0.7501 ms 73.6%
  triton_mm_3042 0.7605 ms 72.6%
  triton_mm_3038 0.7653 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 9.7805 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3025 0.5495 ms 100.0%
  triton_mm_3026 0.5609 ms 98.0%
  triton_mm_3027 0.5772 ms 95.2%
  mm 0.5863 ms 93.7%
  triton_mm_3020 0.7035 ms 78.1%
  triton_mm_3018 0.7273 ms 75.6%
  triton_mm_3021 0.7300 ms 75.3%
  triton_mm_3022 0.7343 ms 74.8%
  triton_mm_3019 0.7621 ms 72.1%
  triton_mm_3023 0.8011 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 9.9808 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4198 ms 100.0%
  triton_mm_3120 0.4244 ms 98.9%
  triton_mm_3121 0.4467 ms 94.0%
  triton_mm_3122 0.4687 ms 89.6%
  triton_mm_3117 0.5208 ms 80.6%
  triton_mm_3113 0.5236 ms 80.2%
  triton_mm_3115 0.5330 ms 78.8%
  triton_mm_3116 0.5590 ms 75.1%
  triton_mm_3118 0.5693 ms 73.8%
  triton_mm_3114 0.5704 ms 73.6%
SingleProcess AUTOTUNE benchmarking takes 9.8533 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3102 0.4058 ms 100.0%
  triton_mm_3101 0.4092 ms 99.2%
  mm 0.4435 ms 91.5%
  triton_mm_3103 0.4815 ms 84.3%
  triton_mm_3096 0.5139 ms 79.0%
  triton_mm_3094 0.5344 ms 75.9%
  triton_mm_3097 0.5390 ms 75.3%
  triton_mm_3098 0.5443 ms 74.5%
  triton_mm_3095 0.5635 ms 72.0%
  triton_mm_3099 0.5894 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 9.7979 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1630 ms 100.0%
  triton_mm_3082 0.1855 ms 87.8%
  triton_mm_3077 0.1938 ms 84.1%
  triton_mm_3083 0.1940 ms 84.0%
  triton_mm_3084 0.1955 ms 83.4%
  triton_mm_3075 0.2041 ms 79.9%
  triton_mm_3079 0.2054 ms 79.3%
  triton_mm_3078 0.2128 ms 76.6%
  triton_mm_3080 0.2197 ms 74.2%
  triton_mm_3076 0.2199 ms 74.1%
SingleProcess AUTOTUNE benchmarking takes 9.9199 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1388 ms 100.0%
  triton_mm_3063 0.1412 ms 98.3%
  triton_mm_3064 0.1431 ms 97.0%
  triton_mm_3065 0.1533 ms 90.5%
  triton_mm_3058 0.1750 ms 79.3%
  triton_mm_3056 0.1820 ms 76.3%
  triton_mm_3060 0.1823 ms 76.1%
  triton_mm_3059 0.1903 ms 72.9%
  triton_mm_3057 0.1934 ms 71.7%
  triton_mm_3061 0.2003 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 10.3730 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3152 0.0324 ms 100.0%
  triton_mm_3145 0.0328 ms 98.6%
  triton_mm_3151 0.0328 ms 98.5%
  triton_mm_3148 0.0337 ms 96.0%
  triton_mm_3149 0.0339 ms 95.4%
  triton_mm_3147 0.0341 ms 94.9%
  triton_mm_3150 0.0348 ms 93.1%
  triton_mm_3154 0.0351 ms 92.2%
  triton_mm_3155 0.0351 ms 92.2%
  triton_mm_3146 0.0352 ms 92.0%
SingleProcess AUTOTUNE benchmarking takes 5.3136 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3191 0.5564 ms 100.0%
  mm 0.5790 ms 96.1%
  triton_mm_3192 0.5853 ms 95.1%
  triton_mm_3193 0.6168 ms 90.2%
  triton_mm_3188 0.6960 ms 79.9%
  triton_mm_3184 0.7089 ms 78.5%
  triton_mm_3186 0.7303 ms 76.2%
  triton_mm_3189 0.7636 ms 72.9%
  triton_mm_3185 0.7719 ms 72.1%
  triton_mm_3187 0.7721 ms 72.1%
SingleProcess AUTOTUNE benchmarking takes 2.6217 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3127 0.0528 ms 100.0%
  triton_mm_3131 0.0532 ms 99.4%
  mm 0.0533 ms 99.2%
  triton_mm_3135 0.0578 ms 91.4%
  triton_mm_3139 0.0682 ms 77.5%
  triton_mm_3126 0.0848 ms 62.3%
  triton_mm_3125 0.0884 ms 59.8%
  triton_mm_3130 0.0953 ms 55.5%
  triton_mm_3134 0.1017 ms 52.0%
  triton_mm_3124 0.1032 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 6.5954 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  mm 0.5517 ms 100.0%
  triton_mm_3172 0.5604 ms 98.5%
  triton_mm_3173 0.5682 ms 97.1%
  triton_mm_3174 0.6200 ms 89.0%
  triton_mm_3169 0.7110 ms 77.6%
  triton_mm_3167 0.7140 ms 77.3%
  triton_mm_3165 0.7184 ms 76.8%
  triton_mm_3168 0.7732 ms 71.4%
  triton_mm_3166 0.8056 ms 68.5%
  triton_mm_3170 0.8274 ms 66.7%
SingleProcess AUTOTUNE benchmarking takes 2.6221 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5476 ms 100.0%
  triton_mm_3229 0.5575 ms 98.2%
  triton_mm_3230 0.5877 ms 93.2%
  triton_mm_3231 0.6225 ms 88.0%
  triton_mm_3222 0.6977 ms 78.5%
  triton_mm_3226 0.7045 ms 77.7%
  triton_mm_3224 0.7189 ms 76.2%
  triton_mm_3225 0.7575 ms 72.3%
  triton_mm_3227 0.7695 ms 71.2%
  triton_mm_3223 0.7752 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.6149 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3210 0.5522 ms 100.0%
  triton_mm_3211 0.5639 ms 97.9%
  triton_mm_3212 0.5795 ms 95.3%
  mm 0.5900 ms 93.6%
  triton_mm_3205 0.7048 ms 78.4%
  triton_mm_3203 0.7352 ms 75.1%
  triton_mm_3206 0.7392 ms 74.7%
  triton_mm_3207 0.7399 ms 74.6%
  triton_mm_3204 0.7758 ms 71.2%
  triton_mm_3208 0.7965 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.6310 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4164 ms 100.0%
  triton_mm_3305 0.4297 ms 96.9%
  triton_mm_3306 0.4525 ms 92.0%
  triton_mm_3307 0.4757 ms 87.5%
  triton_mm_3302 0.5308 ms 78.5%
  triton_mm_3298 0.5370 ms 77.6%
  triton_mm_3300 0.5435 ms 76.6%
  triton_mm_3301 0.5725 ms 72.7%
  triton_mm_3303 0.5806 ms 71.7%
  triton_mm_3299 0.5830 ms 71.4%
SingleProcess AUTOTUNE benchmarking takes 2.5861 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3287 0.4123 ms 100.0%
  triton_mm_3286 0.4129 ms 99.9%
  triton_mm_3288 0.4370 ms 94.4%
  mm 0.4387 ms 94.0%
  triton_mm_3281 0.5237 ms 78.7%
  triton_mm_3279 0.5469 ms 75.4%
  triton_mm_3283 0.5511 ms 74.8%
  triton_mm_3282 0.5538 ms 74.4%
  triton_mm_3280 0.5791 ms 71.2%
  triton_mm_3284 0.5997 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.5935 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1642 ms 100.0%
  triton_mm_3267 0.1862 ms 88.2%
  triton_mm_3268 0.1945 ms 84.4%
  triton_mm_3262 0.1959 ms 83.8%
  triton_mm_3269 0.1971 ms 83.3%
  triton_mm_3264 0.2054 ms 79.9%
  triton_mm_3260 0.2062 ms 79.6%
  triton_mm_3263 0.2136 ms 76.9%
  triton_mm_3265 0.2206 ms 74.4%
  triton_mm_3261 0.2208 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 2.4830 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1402 ms 100.0%
  triton_mm_3248 0.1425 ms 98.4%
  triton_mm_3249 0.1437 ms 97.6%
  triton_mm_3250 0.1550 ms 90.5%
  triton_mm_3243 0.1781 ms 78.7%
  triton_mm_3241 0.1837 ms 76.3%
  triton_mm_3245 0.1853 ms 75.7%
  triton_mm_3244 0.1927 ms 72.8%
  triton_mm_3242 0.1957 ms 71.6%
  triton_mm_3246 0.2020 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.4800 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3334 0.0328 ms 100.0%
  triton_mm_3337 0.0335 ms 97.7%
  triton_mm_3339 0.0337 ms 97.3%
  triton_mm_3333 0.0337 ms 97.2%
  triton_mm_3336 0.0339 ms 96.8%
  triton_mm_3330 0.0340 ms 96.4%
  triton_mm_3332 0.0340 ms 96.3%
  triton_mm_3335 0.0347 ms 94.6%
  triton_mm_3331 0.0350 ms 93.5%
  triton_mm_3338 0.0362 ms 90.5%
SingleProcess AUTOTUNE benchmarking takes 2.0273 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3376 0.5584 ms 100.0%
  mm 0.5738 ms 97.3%
  triton_mm_3377 0.5849 ms 95.5%
  triton_mm_3378 0.6140 ms 90.9%
  triton_mm_3373 0.6991 ms 79.9%
  triton_mm_3369 0.7221 ms 77.3%
  triton_mm_3371 0.7343 ms 76.0%
  triton_mm_3374 0.7599 ms 73.5%
  triton_mm_3372 0.7697 ms 72.6%
  triton_mm_3370 0.7808 ms 71.5%
SingleProcess AUTOTUNE benchmarking takes 2.6223 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3316 0.0532 ms 100.0%
  mm 0.0532 ms 99.9%
  triton_mm_3312 0.0536 ms 99.2%
  triton_mm_3320 0.0588 ms 90.5%
  triton_mm_3324 0.0681 ms 78.0%
  triton_mm_3311 0.0839 ms 63.4%
  triton_mm_3310 0.0884 ms 60.2%
  triton_mm_3315 0.0944 ms 56.3%
  triton_mm_3319 0.1016 ms 52.4%
  triton_mm_3309 0.1039 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 2.1486 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  mm 0.5605 ms 100.0%
  triton_mm_3357 0.5634 ms 99.5%
  triton_mm_3358 0.5688 ms 98.5%
  triton_mm_3359 0.6219 ms 90.1%
  triton_mm_3352 0.7075 ms 79.2%
  triton_mm_3354 0.7172 ms 78.2%
  triton_mm_3350 0.7175 ms 78.1%
  triton_mm_3353 0.7719 ms 72.6%
  triton_mm_3351 0.8072 ms 69.4%
  triton_mm_3355 0.8271 ms 67.8%
SingleProcess AUTOTUNE benchmarking takes 2.6183 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5471 ms 100.0%
  triton_mm_3414 0.5566 ms 98.3%
  triton_mm_3415 0.5873 ms 93.2%
  triton_mm_3416 0.6205 ms 88.2%
  triton_mm_3407 0.7069 ms 77.4%
  triton_mm_3411 0.7153 ms 76.5%
  triton_mm_3409 0.7196 ms 76.0%
  triton_mm_3410 0.7593 ms 72.0%
  triton_mm_3408 0.7718 ms 70.9%
  triton_mm_3412 0.7732 ms 70.8%
SingleProcess AUTOTUNE benchmarking takes 2.6215 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3395 0.5523 ms 100.0%
  triton_mm_3396 0.5655 ms 97.7%
  triton_mm_3397 0.5784 ms 95.5%
  mm 0.5881 ms 93.9%
  triton_mm_3390 0.7063 ms 78.2%
  triton_mm_3388 0.7305 ms 75.6%
  triton_mm_3391 0.7328 ms 75.4%
  triton_mm_3392 0.7409 ms 74.5%
  triton_mm_3389 0.7720 ms 71.5%
  triton_mm_3393 0.8026 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.6288 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1388 ms 100.0%
  triton_mm_3433 0.1420 ms 97.8%
  triton_mm_3434 0.1432 ms 96.9%
  triton_mm_3435 0.1539 ms 90.2%
  triton_mm_3428 0.1762 ms 78.8%
  triton_mm_3426 0.1846 ms 75.2%
  triton_mm_3430 0.1859 ms 74.7%
  triton_mm_3429 0.1917 ms 72.4%
  triton_mm_3427 0.1939 ms 71.6%
  triton_mm_3431 0.2004 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.4807 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3542 0.5611 ms 100.0%
  mm 0.5619 ms 99.9%
  triton_mm_3543 0.5682 ms 98.8%
  triton_mm_3544 0.6180 ms 90.8%
  triton_mm_3537 0.7179 ms 78.2%
  triton_mm_3539 0.7220 ms 77.7%
  triton_mm_3535 0.7245 ms 77.4%
  triton_mm_3538 0.7726 ms 72.6%
  triton_mm_3536 0.8064 ms 69.6%
  triton_mm_3540 0.8266 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 2.6216 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5464 ms 100.0%
  triton_mm_3599 0.5574 ms 98.0%
  triton_mm_3600 0.5884 ms 92.9%
  triton_mm_3601 0.6158 ms 88.7%
  triton_mm_3592 0.7038 ms 77.6%
  triton_mm_3596 0.7058 ms 77.4%
  triton_mm_3594 0.7186 ms 76.0%
  triton_mm_3595 0.7653 ms 71.4%
  triton_mm_3593 0.7710 ms 70.9%
  triton_mm_3597 0.7726 ms 70.7%
SingleProcess AUTOTUNE benchmarking takes 2.6168 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3580 0.5488 ms 100.0%
  triton_mm_3581 0.5706 ms 96.2%
  triton_mm_3582 0.5798 ms 94.6%
  mm 0.5883 ms 93.3%
  triton_mm_3575 0.7125 ms 77.0%
  triton_mm_3573 0.7268 ms 75.5%
  triton_mm_3577 0.7394 ms 74.2%
  triton_mm_3576 0.7405 ms 74.1%
  triton_mm_3574 0.7751 ms 70.8%
  triton_mm_3578 0.7965 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 2.6308 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3727 0.5544 ms 100.0%
  mm 0.5618 ms 98.7%
  triton_mm_3728 0.5676 ms 97.7%
  triton_mm_3729 0.6193 ms 89.5%
  triton_mm_3722 0.7122 ms 77.8%
  triton_mm_3724 0.7126 ms 77.8%
  triton_mm_3720 0.7163 ms 77.4%
  triton_mm_3723 0.7699 ms 72.0%
  triton_mm_3721 0.7965 ms 69.6%
  triton_mm_3725 0.8264 ms 67.1%
SingleProcess AUTOTUNE benchmarking takes 2.6232 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5511 ms 100.0%
  triton_mm_3784 0.5565 ms 99.0%
  triton_mm_3785 0.5871 ms 93.9%
  triton_mm_3786 0.6154 ms 89.6%
  triton_mm_3777 0.6970 ms 79.1%
  triton_mm_3781 0.6983 ms 78.9%
  triton_mm_3779 0.7197 ms 76.6%
  triton_mm_3780 0.7566 ms 72.8%
  triton_mm_3782 0.7647 ms 72.1%
  triton_mm_3778 0.7709 ms 71.5%
SingleProcess AUTOTUNE benchmarking takes 2.6266 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3765 0.5476 ms 100.0%
  triton_mm_3766 0.5642 ms 97.1%
  triton_mm_3767 0.5784 ms 94.7%
  mm 0.5854 ms 93.5%
  triton_mm_3760 0.7048 ms 77.7%
  triton_mm_3758 0.7272 ms 75.3%
  triton_mm_3762 0.7348 ms 74.5%
  triton_mm_3761 0.7379 ms 74.2%
  triton_mm_3759 0.7707 ms 71.1%
  triton_mm_3763 0.8008 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.6367 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5512 ms 100.0%
  triton_mm_3969 0.5526 ms 99.7%
  triton_mm_3970 0.5862 ms 94.0%
  triton_mm_3971 0.6150 ms 89.6%
  triton_mm_3962 0.6944 ms 79.4%
  triton_mm_3966 0.7093 ms 77.7%
  triton_mm_3964 0.7165 ms 76.9%
  triton_mm_3967 0.7623 ms 72.3%
  triton_mm_3963 0.7632 ms 72.2%
  triton_mm_3965 0.7633 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 2.6271 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3950 0.5468 ms 100.0%
  triton_mm_3951 0.5639 ms 97.0%
  triton_mm_3952 0.5781 ms 94.6%
  mm 0.5854 ms 93.4%
  triton_mm_3945 0.7005 ms 78.1%
  triton_mm_3943 0.7265 ms 75.3%
  triton_mm_3946 0.7326 ms 74.6%
  triton_mm_3947 0.7337 ms 74.5%
  triton_mm_3944 0.7646 ms 71.5%
  triton_mm_3948 0.7927 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.6377 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_4135 0.5520 ms 100.0%
  triton_mm_4136 0.5719 ms 96.5%
  triton_mm_4137 0.5839 ms 94.5%
  mm 0.5872 ms 94.0%
  triton_mm_4130 0.6981 ms 79.1%
  triton_mm_4128 0.7281 ms 75.8%
  triton_mm_4132 0.7282 ms 75.8%
  triton_mm_4131 0.7314 ms 75.5%
  triton_mm_4129 0.7620 ms 72.4%
  triton_mm_4133 0.8003 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.6400 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x768, 768x2048)
  triton_mm_7427 0.0746 ms 100.0%
  mm 0.0757 ms 98.6%
  triton_mm_7428 0.0770 ms 96.8%
  triton_mm_7424 0.0819 ms 91.1%
  triton_mm_7420 0.0826 ms 90.3%
  triton_mm_7422 0.0827 ms 90.2%
  triton_mm_7421 0.0894 ms 83.5%
  triton_mm_7425 0.0936 ms 79.7%
  triton_mm_7429 0.0946 ms 78.9%
  triton_mm_7423 0.1080 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.5598 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x2048)
  mm 0.1712 ms 100.0%
  triton_mm_7465 0.1716 ms 99.8%
  triton_mm_7466 0.1798 ms 95.2%
  triton_mm_7467 0.1878 ms 91.1%
  triton_mm_7462 0.1976 ms 86.6%
  triton_mm_7458 0.2003 ms 85.5%
  triton_mm_7460 0.2004 ms 85.4%
  triton_mm_7461 0.2166 ms 79.0%
  triton_mm_7459 0.2229 ms 76.8%
  triton_mm_7463 0.2291 ms 74.7%
SingleProcess AUTOTUNE benchmarking takes 10.3582 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x2048)
  mm 0.1672 ms 100.0%
  triton_mm_7484 0.1894 ms 88.2%
  triton_mm_7485 0.1952 ms 85.6%
  triton_mm_7486 0.1993 ms 83.9%
  triton_mm_7481 0.2098 ms 79.7%
  triton_mm_7477 0.2116 ms 79.0%
  triton_mm_7480 0.2184 ms 76.6%
  triton_mm_7479 0.2184 ms 76.5%
  triton_mm_7482 0.2204 ms 75.9%
  triton_mm_7478 0.2245 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 9.9106 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x1920)
  mm 0.1639 ms 100.0%
  triton_mm_7522 0.1877 ms 87.3%
  triton_mm_7523 0.1937 ms 84.6%
  triton_mm_7517 0.2031 ms 80.7%
  triton_mm_7515 0.2056 ms 79.7%
  triton_mm_7519 0.2073 ms 79.1%
  triton_mm_7518 0.2136 ms 76.7%
  triton_mm_7516 0.2211 ms 74.1%
  triton_mm_7520 0.2224 ms 73.7%
  triton_mm_7524 0.2366 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 10.0178 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x4576, 4576x2048)
  mm 0.0686 ms 100.0%
  triton_mm_7447 0.0774 ms 88.6%
  triton_mm_7441 0.0822 ms 83.4%
  triton_mm_7448 0.0859 ms 79.9%
  triton_mm_7446 0.0864 ms 79.4%
  triton_mm_7443 0.0867 ms 79.1%
  triton_mm_7439 0.0868 ms 79.0%
  triton_mm_7442 0.0898 ms 76.3%
  triton_mm_7444 0.0902 ms 76.0%
  triton_mm_7440 0.0925 ms 74.2%
SingleProcess AUTOTUNE benchmarking takes 9.7985 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x1920)
  mm 0.1492 ms 100.0%
  triton_mm_7503 0.1509 ms 98.8%
  triton_mm_7504 0.1565 ms 95.3%
  triton_mm_7505 0.1609 ms 92.7%
  triton_mm_7498 0.1885 ms 79.1%
  triton_mm_7496 0.1931 ms 77.2%
  triton_mm_7500 0.1932 ms 77.2%
  triton_mm_7497 0.2034 ms 73.3%
  triton_mm_7499 0.2041 ms 73.1%
  triton_mm_7501 0.2094 ms 71.2%
SingleProcess AUTOTUNE benchmarking takes 10.3363 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4560, 4560x32)
  triton_mm_8639 0.0311 ms 100.0%
  triton_mm_8646 0.0316 ms 98.2%
  mm 0.0334 ms 93.1%
  triton_mm_8637 0.0417 ms 74.5%
  triton_mm_8651 0.0437 ms 71.1%
  triton_mm_8638 0.0468 ms 66.3%
  triton_mm_8645 0.0472 ms 65.8%
  triton_mm_8636 0.0480 ms 64.7%
  triton_mm_8642 0.0496 ms 62.6%
  triton_mm_8650 0.0521 ms 59.7%
SingleProcess AUTOTUNE benchmarking takes 2.0928 seconds and 0.0907 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(4560x28, 28x1920)
  triton_mm_9 0.0298 ms 100.0%
  triton_mm_11 0.0298 ms 100.0%
  triton_mm_12 0.0298 ms 99.9%
  triton_mm_8 0.0299 ms 99.6%
  triton_mm_7 0.0311 ms 95.6%
  triton_mm_6 0.0312 ms 95.5%
  triton_mm_15 0.0324 ms 91.8%
  triton_mm_4 0.0328 ms 90.6%
  triton_mm_3 0.0335 ms 88.9%
  triton_mm_14 0.0343 ms 86.8%
SingleProcess AUTOTUNE benchmarking takes 1.9727 seconds and 0.0088 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_29 0.0466 ms 100.0%
  triton_mm_25 0.0476 ms 98.0%
  triton_mm_28 0.0487 ms 95.7%
  triton_mm_33 0.0488 ms 95.6%
  triton_mm_18 0.0499 ms 93.5%
  triton_mm_24 0.0502 ms 92.9%
  triton_mm_21 0.0505 ms 92.3%
  triton_mm_19 0.0512 ms 91.1%
  bias_addmm 0.0516 ms 90.4%
  triton_mm_31 0.0533 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.2372 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_51 0.4383 ms 100.0%
  triton_mm_50 0.4460 ms 98.3%
  triton_mm_52 0.4840 ms 90.5%
  bias_addmm 0.5228 ms 83.8%
  triton_mm_45 0.5387 ms 81.4%
  triton_mm_47 0.5647 ms 77.6%
  triton_mm_43 0.5726 ms 76.5%
  triton_mm_46 0.5952 ms 73.6%
  triton_mm_48 0.6346 ms 69.1%
  triton_mm_44 0.6510 ms 67.3%
SingleProcess AUTOTUNE benchmarking takes 2.7238 seconds and 0.0177 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_70 0.1473 ms 100.0%
  triton_mm_69 0.1515 ms 97.2%
  triton_mm_71 0.1604 ms 91.8%
  bias_addmm 0.1628 ms 90.4%
  triton_mm_64 0.1823 ms 80.8%
  addmm 0.1867 ms 78.9%
  triton_mm_62 0.1873 ms 78.6%
  triton_mm_66 0.1950 ms 75.5%
  triton_mm_65 0.2017 ms 73.0%
  triton_mm_67 0.2111 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.5990 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(4576x7680, 4576x1920, 1920x7680)
  triton_mm_89 0.5834 ms 100.0%
  triton_mm_88 0.5918 ms 98.6%
  triton_mm_90 0.6435 ms 90.7%
  bias_addmm 0.6482 ms 90.0%
  triton_mm_83 0.7148 ms 81.6%
  triton_mm_85 0.7415 ms 78.7%
  triton_mm_81 0.7627 ms 76.5%
  addmm 0.7865 ms 74.2%
  triton_mm_84 0.8010 ms 72.8%
  triton_mm_86 0.8397 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.7526 seconds and 0.3181 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_108 0.5525 ms 100.0%
  triton_mm_107 0.5753 ms 96.0%
  triton_mm_109 0.5871 ms 94.1%
  bias_addmm 0.6184 ms 89.3%
  addmm 0.6600 ms 83.7%
  triton_mm_102 0.7050 ms 78.4%
  triton_mm_103 0.7321 ms 75.5%
  triton_mm_100 0.7428 ms 74.4%
  triton_mm_104 0.7451 ms 74.2%
  triton_mm_105 0.7986 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7528 seconds and 0.2460 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_122 0.0466 ms 100.0%
  triton_mm_118 0.0474 ms 98.3%
  triton_mm_126 0.0486 ms 95.8%
  triton_mm_121 0.0491 ms 95.0%
  triton_mm_117 0.0491 ms 94.8%
  triton_mm_111 0.0496 ms 93.9%
  triton_mm_114 0.0512 ms 91.1%
  bias_addmm 0.0515 ms 90.5%
  triton_mm_112 0.0517 ms 90.2%
  triton_mm_113 0.0527 ms 88.4%
SingleProcess AUTOTUNE benchmarking takes 2.2169 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_201 0.5518 ms 100.0%
  triton_mm_200 0.5746 ms 96.0%
  triton_mm_202 0.5834 ms 94.6%
  bias_addmm 0.6198 ms 89.0%
  addmm 0.6273 ms 88.0%
  triton_mm_195 0.7053 ms 78.2%
  triton_mm_196 0.7307 ms 75.5%
  triton_mm_193 0.7426 ms 74.3%
  triton_mm_197 0.7449 ms 74.1%
  triton_mm_198 0.7944 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.7403 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_215 0.0474 ms 100.0%
  triton_mm_211 0.0482 ms 98.5%
  triton_mm_219 0.0487 ms 97.4%
  triton_mm_214 0.0490 ms 96.9%
  triton_mm_210 0.0491 ms 96.7%
  triton_mm_204 0.0495 ms 95.7%
  triton_mm_205 0.0509 ms 93.1%
  triton_mm_207 0.0512 ms 92.7%
  bias_addmm 0.0522 ms 90.9%
  triton_mm_206 0.0527 ms 90.0%
SingleProcess AUTOTUNE benchmarking takes 2.2203 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_294 0.5512 ms 100.0%
  triton_mm_293 0.5739 ms 96.0%
  triton_mm_295 0.5824 ms 94.6%
  bias_addmm 0.6204 ms 88.9%
  addmm 0.6318 ms 87.3%
  triton_mm_288 0.7040 ms 78.3%
  triton_mm_289 0.7312 ms 75.4%
  triton_mm_286 0.7417 ms 74.3%
  triton_mm_290 0.7456 ms 73.9%
  triton_mm_291 0.7941 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.7510 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_308 0.0466 ms 100.0%
  triton_mm_304 0.0482 ms 96.7%
  triton_mm_312 0.0487 ms 95.7%
  triton_mm_303 0.0490 ms 95.2%
  triton_mm_307 0.0501 ms 93.1%
  triton_mm_297 0.0502 ms 92.9%
  triton_mm_298 0.0508 ms 91.7%
  triton_mm_300 0.0513 ms 90.8%
  bias_addmm 0.0522 ms 89.4%
  triton_mm_299 0.0527 ms 88.5%
SingleProcess AUTOTUNE benchmarking takes 2.2230 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_401 0.0466 ms 100.0%
  triton_mm_405 0.0479 ms 97.3%
  triton_mm_397 0.0482 ms 96.8%
  triton_mm_396 0.0490 ms 95.2%
  triton_mm_390 0.0496 ms 94.1%
  triton_mm_400 0.0501 ms 93.1%
  triton_mm_393 0.0504 ms 92.6%
  triton_mm_391 0.0510 ms 91.5%
  bias_addmm 0.0517 ms 90.1%
  triton_mm_392 0.0533 ms 87.4%
SingleProcess AUTOTUNE benchmarking takes 2.2244 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_490 0.0473 ms 100.0%
  triton_mm_494 0.0474 ms 99.9%
  triton_mm_498 0.0486 ms 97.3%
  triton_mm_493 0.0490 ms 96.6%
  triton_mm_489 0.0492 ms 96.2%
  triton_mm_486 0.0504 ms 94.0%
  triton_mm_483 0.0504 ms 93.8%
  triton_mm_484 0.0513 ms 92.2%
  bias_addmm 0.0517 ms 91.5%
  triton_mm_485 0.0533 ms 88.7%
SingleProcess AUTOTUNE benchmarking takes 2.2186 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x1920, 1920x2048)
  triton_mm_591 0.1698 ms 100.0%
  triton_mm_593 0.1844 ms 92.1%
  triton_mm_586 0.1903 ms 89.2%
  triton_mm_584 0.1996 ms 85.1%
  addmm 0.2013 ms 84.3%
  triton_mm_592 0.2015 ms 84.3%
  bias_addmm 0.2039 ms 83.3%
  triton_mm_587 0.2128 ms 79.8%
  triton_mm_589 0.2278 ms 74.5%
  triton_mm_585 0.2300 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 11.2091 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x2048, 2048x2048)
  triton_mm_611 0.1772 ms 100.0%
  triton_mm_610 0.1778 ms 99.7%
  bias_addmm 0.1802 ms 98.4%
  triton_mm_612 0.1932 ms 91.7%
  triton_mm_605 0.2040 ms 86.9%
  triton_mm_603 0.2046 ms 86.6%
  triton_mm_607 0.2059 ms 86.1%
  addmm 0.2082 ms 85.1%
  triton_mm_606 0.2229 ms 79.5%
  triton_mm_608 0.2397 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 2.8744 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x768)
  mm 0.0635 ms 100.0%
  triton_mm_629 0.0646 ms 98.2%
  triton_mm_630 0.0653 ms 97.2%
  triton_mm_631 0.0840 ms 75.5%
  triton_mm_625 0.0883 ms 71.9%
  triton_mm_627 0.0917 ms 69.2%
  triton_mm_623 0.0921 ms 68.9%
  triton_mm_626 0.0925 ms 68.6%
  triton_mm_622 0.0935 ms 67.9%
  triton_mm_624 0.0986 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 10.1887 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_644 0.0466 ms 100.0%
  triton_mm_648 0.0479 ms 97.2%
  triton_mm_640 0.0484 ms 96.4%
  triton_mm_633 0.0497 ms 93.8%
  triton_mm_639 0.0501 ms 93.1%
  triton_mm_643 0.0501 ms 93.1%
  bias_addmm 0.0506 ms 92.1%
  triton_mm_636 0.0512 ms 91.0%
  triton_mm_634 0.0516 ms 90.4%
  triton_mm_635 0.0528 ms 88.3%
SingleProcess AUTOTUNE benchmarking takes 2.2284 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_733 0.0474 ms 100.0%
  triton_mm_737 0.0474 ms 99.9%
  triton_mm_741 0.0479 ms 98.9%
  triton_mm_732 0.0489 ms 96.9%
  triton_mm_736 0.0490 ms 96.7%
  triton_mm_726 0.0498 ms 95.2%
  bias_addmm 0.0507 ms 93.6%
  triton_mm_727 0.0509 ms 93.0%
  triton_mm_729 0.0512 ms 92.6%
  triton_mm_728 0.0532 ms 89.2%
SingleProcess AUTOTUNE benchmarking takes 2.5512 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_830 0.0475 ms 100.0%
  triton_mm_834 0.0479 ms 99.1%
  triton_mm_826 0.0481 ms 98.7%
  triton_mm_829 0.0490 ms 96.8%
  triton_mm_825 0.0492 ms 96.4%
  triton_mm_819 0.0497 ms 95.6%
  triton_mm_820 0.0509 ms 93.3%
  triton_mm_822 0.0511 ms 92.9%
  bias_addmm 0.0519 ms 91.5%
  triton_mm_821 0.0533 ms 89.0%
SingleProcess AUTOTUNE benchmarking takes 2.2441 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_923 0.0466 ms 100.0%
  triton_mm_927 0.0479 ms 97.3%
  triton_mm_919 0.0481 ms 96.8%
  triton_mm_918 0.0491 ms 94.9%
  triton_mm_922 0.0492 ms 94.7%
  triton_mm_912 0.0496 ms 94.0%
  triton_mm_913 0.0510 ms 91.4%
  triton_mm_915 0.0512 ms 91.1%
  bias_addmm 0.0523 ms 89.2%
  triton_mm_914 0.0528 ms 88.3%
SingleProcess AUTOTUNE benchmarking takes 2.2431 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_1016 0.0467 ms 100.0%
  triton_mm_1020 0.0479 ms 97.4%
  triton_mm_1012 0.0481 ms 97.0%
  triton_mm_1005 0.0497 ms 93.8%
  triton_mm_1011 0.0500 ms 93.3%
  triton_mm_1015 0.0501 ms 93.1%
  triton_mm_1006 0.0510 ms 91.5%
  triton_mm_1008 0.0512 ms 91.2%
  bias_addmm 0.0525 ms 88.9%
  triton_mm_1007 0.0528 ms 88.4%
SingleProcess AUTOTUNE benchmarking takes 2.2278 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x3840, 16x1920, 1920x3840)
  triton_mm_2876 0.0216 ms 100.0%
  triton_mm_2880 0.0228 ms 94.4%
  triton_mm_2872 0.0229 ms 94.2%
  triton_mm_2868 0.0235 ms 91.8%
  bias_addmm 0.0252 ms 85.5%
  triton_mm_2865 0.0270 ms 79.9%
  addmm 0.0275 ms 78.5%
  triton_mm_2871 0.0275 ms 78.5%
  triton_mm_2875 0.0275 ms 78.4%
  triton_mm_2867 0.0276 ms 78.2%
SingleProcess AUTOTUNE benchmarking takes 7.6790 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x32768, 4576x1920, 1920x32768)
  triton_mm_2898 2.4654 ms 100.0%
  triton_mm_2897 2.4904 ms 99.0%
  bias_addmm 2.6103 ms 94.4%
  triton_mm_2899 2.7205 ms 90.6%
  addmm 2.8868 ms 85.4%
  triton_mm_2892 3.0306 ms 81.3%
  triton_mm_2894 3.1485 ms 78.3%
  triton_mm_2890 3.2242 ms 76.5%
  triton_mm_2893 3.2838 ms 75.1%
  triton_mm_2895 3.5793 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 12.0097 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x4576, 4576x1920)
  triton_mm_2935 2.2651 ms 100.0%
  mm 2.2906 ms 98.9%
  triton_mm_2936 2.4012 ms 94.3%
  triton_mm_2937 2.7233 ms 83.2%
  triton_mm_2928 2.9690 ms 76.3%
  triton_mm_2932 3.0279 ms 74.8%
  triton_mm_2930 3.0676 ms 73.8%
  triton_mm_2931 3.1383 ms 72.2%
  triton_mm_2933 3.1868 ms 71.1%
  triton_mm_2929 3.2708 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 9.8292 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x32768, 32768x1920)
  mm 2.2883 ms 100.0%
  triton_mm_2916 2.3494 ms 97.4%
  triton_mm_2917 2.3867 ms 95.9%
  triton_mm_2918 2.4115 ms 94.9%
  triton_mm_2912 3.0252 ms 75.6%
  triton_mm_2910 3.2405 ms 70.6%
  triton_mm_2909 3.2535 ms 70.3%
  triton_mm_2911 3.2577 ms 70.2%
  triton_mm_2914 3.3698 ms 67.9%
  triton_mm_2913 3.4401 ms 66.5%
SingleProcess AUTOTUNE benchmarking takes 10.0493 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3006 0.5529 ms 100.0%
  mm 0.5756 ms 96.1%
  triton_mm_3007 0.5781 ms 95.6%
  triton_mm_3008 0.6069 ms 91.1%
  triton_mm_3003 0.6822 ms 81.0%
  triton_mm_2999 0.6989 ms 79.1%
  triton_mm_3001 0.7121 ms 77.6%
  triton_mm_3002 0.7443 ms 74.3%
  triton_mm_3004 0.7472 ms 74.0%
  triton_mm_3000 0.7619 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 10.1471 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x16, 16x1920)
  triton_mm_2960 0.0166 ms 100.0%
  triton_mm_2962 0.0166 ms 99.8%
  triton_mm_2964 0.0166 ms 99.8%
  triton_mm_2965 0.0169 ms 97.9%
  triton_mm_2967 0.0171 ms 97.0%
  triton_mm_2966 0.0173 ms 95.6%
  triton_mm_2961 0.0181 ms 91.4%
  triton_mm_2963 0.0182 ms 91.3%
  triton_mm_2969 0.0184 ms 90.2%
  triton_mm_2958 0.0189 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 5.3128 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x3840, 3840x1920)
  triton_mm_2946 0.0217 ms 100.0%
  triton_mm_2950 0.0247 ms 87.7%
  triton_mm_2942 0.0253 ms 85.5%
  mm 0.0283 ms 76.7%
  triton_mm_2954 0.0286 ms 75.8%
  triton_mm_2941 0.0346 ms 62.7%
  triton_mm_2940 0.0351 ms 61.8%
  triton_mm_2945 0.0366 ms 59.2%
  triton_mm_2949 0.0388 ms 55.8%
  triton_mm_2939 0.0406 ms 53.3%
SingleProcess AUTOTUNE benchmarking takes 9.2718 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_2987 0.5543 ms 100.0%
  triton_mm_2988 0.5582 ms 99.3%
  mm 0.5595 ms 99.1%
  triton_mm_2989 0.6072 ms 91.3%
  triton_mm_2984 0.7004 ms 79.1%
  triton_mm_2982 0.7040 ms 78.7%
  triton_mm_2980 0.7070 ms 78.4%
  triton_mm_2983 0.7621 ms 72.7%
  triton_mm_2981 0.7837 ms 70.7%
  triton_mm_2985 0.8021 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 10.4114 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  triton_mm_3044 0.5526 ms 100.0%
  mm 0.5528 ms 100.0%
  triton_mm_3045 0.5790 ms 95.4%
  triton_mm_3046 0.6112 ms 90.4%
  triton_mm_3037 0.6857 ms 80.6%
  triton_mm_3041 0.6894 ms 80.1%
  triton_mm_3039 0.7051 ms 78.4%
  triton_mm_3040 0.7363 ms 75.0%
  triton_mm_3042 0.7521 ms 73.5%
  triton_mm_3038 0.7551 ms 73.2%
SingleProcess AUTOTUNE benchmarking takes 9.7966 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3025 0.5445 ms 100.0%
  triton_mm_3026 0.5563 ms 97.9%
  triton_mm_3027 0.5697 ms 95.6%
  mm 0.5920 ms 92.0%
  triton_mm_3020 0.6947 ms 78.4%
  triton_mm_3021 0.7188 ms 75.7%
  triton_mm_3018 0.7244 ms 75.2%
  triton_mm_3022 0.7282 ms 74.8%
  triton_mm_3019 0.7520 ms 72.4%
  triton_mm_3023 0.7853 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 9.8682 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4210 ms 100.0%
  triton_mm_3120 0.4261 ms 98.8%
  triton_mm_3121 0.4469 ms 94.2%
  triton_mm_3122 0.4693 ms 89.7%
  triton_mm_3113 0.5232 ms 80.5%
  triton_mm_3115 0.5352 ms 78.7%
  triton_mm_3117 0.5690 ms 74.0%
  triton_mm_3118 0.5727 ms 73.5%
  triton_mm_3114 0.5737 ms 73.4%
  triton_mm_3116 0.6630 ms 63.5%
SingleProcess AUTOTUNE benchmarking takes 12.7769 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3102 0.4093 ms 100.0%
  triton_mm_3101 0.4102 ms 99.8%
  triton_mm_3103 0.4300 ms 95.2%
  mm 0.4386 ms 93.3%
  triton_mm_3096 0.5168 ms 79.2%
  triton_mm_3094 0.5393 ms 75.9%
  triton_mm_3097 0.5430 ms 75.4%
  triton_mm_3098 0.5456 ms 75.0%
  triton_mm_3095 0.5642 ms 72.5%
  triton_mm_3099 0.5859 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 9.8089 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1627 ms 100.0%
  triton_mm_3082 0.1848 ms 88.0%
  triton_mm_3075 0.1932 ms 84.2%
  triton_mm_3083 0.1937 ms 84.0%
  triton_mm_3077 0.1945 ms 83.7%
  triton_mm_3084 0.1960 ms 83.0%
  triton_mm_3079 0.1961 ms 83.0%
  triton_mm_3078 0.2148 ms 75.8%
  triton_mm_3076 0.2197 ms 74.1%
  triton_mm_3080 0.2202 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 9.6488 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1385 ms 100.0%
  triton_mm_3063 0.1417 ms 97.7%
  triton_mm_3064 0.1422 ms 97.4%
  triton_mm_3065 0.1519 ms 91.2%
  triton_mm_3058 0.1757 ms 78.8%
  triton_mm_3056 0.1808 ms 76.6%
  triton_mm_3060 0.1841 ms 75.2%
  triton_mm_3059 0.1911 ms 72.5%
  triton_mm_3057 0.1942 ms 71.3%
  triton_mm_3061 0.2006 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 10.2722 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3145 0.0329 ms 100.0%
  triton_mm_3147 0.0329 ms 100.0%
  triton_mm_3151 0.0329 ms 99.8%
  triton_mm_3150 0.0339 ms 97.1%
  triton_mm_3152 0.0340 ms 96.6%
  triton_mm_3149 0.0342 ms 96.2%
  triton_mm_3148 0.0352 ms 93.4%
  triton_mm_3155 0.0352 ms 93.4%
  triton_mm_3154 0.0353 ms 93.1%
  triton_mm_3146 0.0354 ms 92.9%
SingleProcess AUTOTUNE benchmarking takes 5.4162 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3191 0.5610 ms 100.0%
  mm 0.5766 ms 97.3%
  triton_mm_3192 0.5855 ms 95.8%
  triton_mm_3193 0.6148 ms 91.3%
  triton_mm_3188 0.6955 ms 80.7%
  triton_mm_3184 0.7139 ms 78.6%
  triton_mm_3186 0.7246 ms 77.4%
  triton_mm_3189 0.7574 ms 74.1%
  triton_mm_3187 0.7620 ms 73.6%
  triton_mm_3185 0.7742 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 2.6106 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0533 ms 100.0%
  triton_mm_3131 0.0534 ms 99.8%
  triton_mm_3127 0.0537 ms 99.2%
  triton_mm_3135 0.0590 ms 90.3%
  triton_mm_3139 0.0683 ms 78.0%
  triton_mm_3126 0.0843 ms 63.2%
  triton_mm_3125 0.0879 ms 60.6%
  triton_mm_3130 0.0957 ms 55.7%
  triton_mm_3134 0.1012 ms 52.6%
  triton_mm_3124 0.1027 ms 51.9%
SingleProcess AUTOTUNE benchmarking takes 6.6354 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  mm 0.5610 ms 100.0%
  triton_mm_3172 0.5611 ms 100.0%
  triton_mm_3173 0.5681 ms 98.8%
  triton_mm_3174 0.6191 ms 90.6%
  triton_mm_3167 0.7051 ms 79.6%
  triton_mm_3169 0.7166 ms 78.3%
  triton_mm_3165 0.7182 ms 78.1%
  triton_mm_3168 0.7683 ms 73.0%
  triton_mm_3166 0.7990 ms 70.2%
  triton_mm_3170 0.8185 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.6155 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5497 ms 100.0%
  triton_mm_3229 0.5607 ms 98.0%
  triton_mm_3230 0.5915 ms 92.9%
  triton_mm_3231 0.6215 ms 88.4%
  triton_mm_3222 0.6971 ms 78.9%
  triton_mm_3226 0.7032 ms 78.2%
  triton_mm_3224 0.7180 ms 76.6%
  triton_mm_3225 0.7530 ms 73.0%
  triton_mm_3227 0.7650 ms 71.9%
  triton_mm_3223 0.7674 ms 71.6%
SingleProcess AUTOTUNE benchmarking takes 2.6156 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3210 0.5533 ms 100.0%
  triton_mm_3211 0.5699 ms 97.1%
  triton_mm_3212 0.5795 ms 95.5%
  mm 0.5927 ms 93.4%
  triton_mm_3205 0.7097 ms 78.0%
  triton_mm_3203 0.7323 ms 75.6%
  triton_mm_3206 0.7336 ms 75.4%
  triton_mm_3207 0.7394 ms 74.8%
  triton_mm_3204 0.7684 ms 72.0%
  triton_mm_3208 0.7987 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.6258 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  triton_mm_3305 0.4316 ms 100.0%
  triton_mm_3306 0.4552 ms 94.8%
  mm 0.4728 ms 91.3%
  triton_mm_3307 0.4794 ms 90.0%
  triton_mm_3298 0.5335 ms 80.9%
  triton_mm_3302 0.5353 ms 80.6%
  triton_mm_3300 0.5452 ms 79.2%
  triton_mm_3301 0.5710 ms 75.6%
  triton_mm_3299 0.5834 ms 74.0%
  triton_mm_3303 0.5840 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 2.5982 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3287 0.4158 ms 100.0%
  triton_mm_3286 0.4163 ms 99.9%
  triton_mm_3288 0.4375 ms 95.0%
  mm 0.4421 ms 94.1%
  triton_mm_3281 0.5270 ms 78.9%
  triton_mm_3279 0.5505 ms 75.5%
  triton_mm_3282 0.5540 ms 75.1%
  triton_mm_3283 0.5555 ms 74.9%
  triton_mm_3280 0.5725 ms 72.6%
  triton_mm_3284 0.5964 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.5897 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1638 ms 100.0%
  triton_mm_3267 0.1861 ms 88.0%
  triton_mm_3268 0.1951 ms 84.0%
  triton_mm_3260 0.1953 ms 83.8%
  triton_mm_3262 0.1966 ms 83.3%
  triton_mm_3269 0.1976 ms 82.9%
  triton_mm_3264 0.1980 ms 82.7%
  triton_mm_3263 0.2167 ms 75.6%
  triton_mm_3261 0.2230 ms 73.4%
  triton_mm_3265 0.2237 ms 73.2%
SingleProcess AUTOTUNE benchmarking takes 2.4911 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1399 ms 100.0%
  triton_mm_3248 0.1440 ms 97.2%
  triton_mm_3249 0.1453 ms 96.3%
  triton_mm_3250 0.1552 ms 90.1%
  triton_mm_3243 0.1788 ms 78.2%
  triton_mm_3241 0.1835 ms 76.2%
  triton_mm_3245 0.1866 ms 75.0%
  triton_mm_3244 0.1934 ms 72.3%
  triton_mm_3242 0.1973 ms 70.9%
  triton_mm_3246 0.2045 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.4874 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3337 0.0325 ms 100.0%
  triton_mm_3330 0.0328 ms 98.9%
  triton_mm_3334 0.0329 ms 98.8%
  triton_mm_3336 0.0329 ms 98.7%
  triton_mm_3331 0.0340 ms 95.7%
  triton_mm_3332 0.0341 ms 95.3%
  triton_mm_3333 0.0349 ms 93.1%
  triton_mm_3335 0.0350 ms 92.9%
  triton_mm_3339 0.0351 ms 92.4%
  triton_mm_3338 0.0364 ms 89.2%
SingleProcess AUTOTUNE benchmarking takes 2.0401 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3376 0.5615 ms 100.0%
  mm 0.5794 ms 96.9%
  triton_mm_3377 0.5889 ms 95.3%
  triton_mm_3378 0.6205 ms 90.5%
  triton_mm_3373 0.6949 ms 80.8%
  triton_mm_3369 0.7179 ms 78.2%
  triton_mm_3371 0.7308 ms 76.8%
  triton_mm_3374 0.7587 ms 74.0%
  triton_mm_3372 0.7624 ms 73.6%
  triton_mm_3370 0.7713 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 2.6139 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3312 0.0531 ms 100.0%
  triton_mm_3316 0.0532 ms 99.7%
  mm 0.0532 ms 99.6%
  triton_mm_3320 0.0588 ms 90.2%
  triton_mm_3324 0.0683 ms 77.7%
  triton_mm_3311 0.0843 ms 62.9%
  triton_mm_3310 0.0886 ms 59.9%
  triton_mm_3315 0.0947 ms 56.1%
  triton_mm_3319 0.1012 ms 52.4%
  triton_mm_3309 0.1035 ms 51.3%
SingleProcess AUTOTUNE benchmarking takes 2.1634 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3357 0.5619 ms 100.0%
  mm 0.5622 ms 99.9%
  triton_mm_3358 0.5700 ms 98.6%
  triton_mm_3359 0.6235 ms 90.1%
  triton_mm_3352 0.7045 ms 79.8%
  triton_mm_3354 0.7168 ms 78.4%
  triton_mm_3350 0.7177 ms 78.3%
  triton_mm_3353 0.7749 ms 72.5%
  triton_mm_3351 0.8047 ms 69.8%
  triton_mm_3355 0.8275 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 2.6210 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5500 ms 100.0%
  triton_mm_3414 0.5626 ms 97.7%
  triton_mm_3415 0.5917 ms 92.9%
  triton_mm_3416 0.6216 ms 88.5%
  triton_mm_3407 0.6973 ms 78.9%
  triton_mm_3411 0.7036 ms 78.2%
  triton_mm_3409 0.7204 ms 76.3%
  triton_mm_3410 0.7503 ms 73.3%
  triton_mm_3412 0.7652 ms 71.9%
  triton_mm_3408 0.7722 ms 71.2%
SingleProcess AUTOTUNE benchmarking takes 2.6179 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3395 0.5535 ms 100.0%
  triton_mm_3396 0.5674 ms 97.5%
  triton_mm_3397 0.5801 ms 95.4%
  mm 0.5923 ms 93.4%
  triton_mm_3390 0.7101 ms 77.9%
  triton_mm_3388 0.7328 ms 75.5%
  triton_mm_3391 0.7335 ms 75.5%
  triton_mm_3392 0.7414 ms 74.7%
  triton_mm_3389 0.7689 ms 72.0%
  triton_mm_3393 0.7986 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.6304 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5495 ms 100.0%
  triton_mm_3599 0.5609 ms 98.0%
  triton_mm_3600 0.5917 ms 92.9%
  triton_mm_3601 0.6215 ms 88.4%
  triton_mm_3596 0.7016 ms 78.3%
  triton_mm_3592 0.7034 ms 78.1%
  triton_mm_3594 0.7189 ms 76.4%
  triton_mm_3595 0.7492 ms 73.3%
  triton_mm_3597 0.7648 ms 71.9%
  triton_mm_3593 0.7721 ms 71.2%
SingleProcess AUTOTUNE benchmarking takes 2.6169 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3580 0.5533 ms 100.0%
  triton_mm_3581 0.5679 ms 97.4%
  triton_mm_3582 0.5803 ms 95.3%
  mm 0.5924 ms 93.4%
  triton_mm_3575 0.7092 ms 78.0%
  triton_mm_3573 0.7321 ms 75.6%
  triton_mm_3576 0.7332 ms 75.5%
  triton_mm_3577 0.7405 ms 74.7%
  triton_mm_3574 0.7634 ms 72.5%
  triton_mm_3578 0.7988 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.6242 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x768, 768x2048)
  triton_mm_7427 0.0735 ms 100.0%
  mm 0.0760 ms 96.8%
  triton_mm_7428 0.0768 ms 95.7%
  triton_mm_7424 0.0820 ms 89.7%
  triton_mm_7420 0.0822 ms 89.5%
  triton_mm_7422 0.0832 ms 88.4%
  triton_mm_7421 0.0899 ms 81.8%
  triton_mm_7425 0.0934 ms 78.8%
  triton_mm_7429 0.0956 ms 76.9%
  triton_mm_7423 0.1083 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 10.4155 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x2048)
  triton_mm_7465 0.1702 ms 100.0%
  mm 0.1702 ms 100.0%
  triton_mm_7466 0.1779 ms 95.6%
  triton_mm_7467 0.1868 ms 91.1%
  triton_mm_7460 0.1992 ms 85.4%
  triton_mm_7458 0.1993 ms 85.4%
  triton_mm_7462 0.2010 ms 84.7%
  triton_mm_7461 0.2165 ms 78.6%
  triton_mm_7459 0.2237 ms 76.1%
  triton_mm_7463 0.2285 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 10.2606 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x2048)
  mm 0.1678 ms 100.0%
  triton_mm_7484 0.1878 ms 89.3%
  triton_mm_7485 0.1940 ms 86.5%
  triton_mm_7486 0.2005 ms 83.7%
  triton_mm_7481 0.2121 ms 79.1%
  triton_mm_7477 0.2127 ms 78.9%
  triton_mm_7480 0.2182 ms 76.9%
  triton_mm_7479 0.2188 ms 76.7%
  triton_mm_7482 0.2229 ms 75.3%
  triton_mm_7478 0.2258 ms 74.3%
SingleProcess AUTOTUNE benchmarking takes 9.7172 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x1920)
  mm 0.1653 ms 100.0%
  triton_mm_7523 0.1936 ms 85.4%
  triton_mm_7519 0.1972 ms 83.8%
  triton_mm_7524 0.1975 ms 83.7%
  triton_mm_7517 0.2035 ms 81.2%
  triton_mm_7515 0.2071 ms 79.8%
  triton_mm_7518 0.2149 ms 76.9%
  triton_mm_7522 0.2198 ms 75.2%
  triton_mm_7516 0.2236 ms 73.9%
  triton_mm_7520 0.2240 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 9.7937 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x4576, 4576x2048)
  mm 0.0697 ms 100.0%
  triton_mm_7448 0.0728 ms 95.7%
  triton_mm_7447 0.0774 ms 90.1%
  triton_mm_7441 0.0829 ms 84.1%
  triton_mm_7443 0.0885 ms 78.8%
  triton_mm_7439 0.0893 ms 78.1%
  triton_mm_7446 0.0893 ms 78.0%
  triton_mm_7442 0.0914 ms 76.3%
  triton_mm_7440 0.0922 ms 75.7%
  triton_mm_7444 0.0923 ms 75.5%
SingleProcess AUTOTUNE benchmarking takes 9.7603 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x1920)
  mm 0.1481 ms 100.0%
  triton_mm_7503 0.1517 ms 97.6%
  triton_mm_7504 0.1583 ms 93.6%
  triton_mm_7505 0.1620 ms 91.5%
  triton_mm_7498 0.1904 ms 77.8%
  triton_mm_7496 0.1940 ms 76.4%
  triton_mm_7500 0.1973 ms 75.1%
  triton_mm_7499 0.2044 ms 72.5%
  triton_mm_7497 0.2070 ms 71.6%
  triton_mm_7501 0.2117 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 10.3150 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4560, 4560x28)
  triton_mm_8646 0.0350 ms 100.0%
  triton_mm_8639 0.0359 ms 97.5%
  mm 0.0366 ms 95.6%
  triton_mm_8637 0.0423 ms 82.9%
  triton_mm_8638 0.0473 ms 74.1%
  triton_mm_8645 0.0501 ms 69.9%
  triton_mm_8651 0.0506 ms 69.3%
  triton_mm_8636 0.0527 ms 66.4%
  triton_mm_8642 0.0535 ms 65.5%
  triton_mm_8650 0.0557 ms 62.9%
SingleProcess AUTOTUNE benchmarking takes 6.8581 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(4560x28, 28x1920)
  triton_mm_7 0.0293 ms 100.0%
  triton_mm_12 0.0294 ms 99.8%
  triton_mm_8 0.0295 ms 99.5%
  triton_mm_6 0.0295 ms 99.5%
  triton_mm_11 0.0295 ms 99.3%
  triton_mm_9 0.0296 ms 99.2%
  triton_mm_15 0.0319 ms 92.1%
  triton_mm_4 0.0322 ms 91.2%
  triton_mm_2 0.0322 ms 91.1%
  triton_mm_14 0.0328 ms 89.6%
SingleProcess AUTOTUNE benchmarking takes 2.0943 seconds and 0.0201 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_29 0.0466 ms 100.0%
  triton_mm_25 0.0475 ms 98.0%
  triton_mm_33 0.0488 ms 95.3%
  triton_mm_24 0.0495 ms 94.1%
  triton_mm_28 0.0501 ms 92.9%
  triton_mm_18 0.0504 ms 92.4%
  bias_addmm 0.0509 ms 91.5%
  triton_mm_21 0.0511 ms 91.2%
  triton_mm_19 0.0514 ms 90.6%
  triton_mm_20 0.0532 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.1999 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_51 0.4331 ms 100.0%
  triton_mm_50 0.4419 ms 98.0%
  triton_mm_52 0.4772 ms 90.8%
  bias_addmm 0.5197 ms 83.3%
  triton_mm_45 0.5366 ms 80.7%
  triton_mm_47 0.5531 ms 78.3%
  triton_mm_43 0.5659 ms 76.5%
  triton_mm_46 0.5906 ms 73.3%
  addmm 0.5944 ms 72.9%
  triton_mm_48 0.6244 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.7490 seconds and 0.0132 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_70 0.1465 ms 100.0%
  triton_mm_69 0.1508 ms 97.1%
  triton_mm_71 0.1592 ms 92.0%
  bias_addmm 0.1613 ms 90.8%
  triton_mm_64 0.1779 ms 82.3%
  addmm 0.1846 ms 79.3%
  triton_mm_62 0.1848 ms 79.3%
  triton_mm_66 0.1902 ms 77.0%
  triton_mm_65 0.2000 ms 73.2%
  triton_mm_67 0.2075 ms 70.6%
SingleProcess AUTOTUNE benchmarking takes 2.5876 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(4576x7680, 4576x1920, 1920x7680)
  triton_mm_89 0.5774 ms 100.0%
  triton_mm_88 0.5883 ms 98.1%
  triton_mm_90 0.6398 ms 90.2%
  bias_addmm 0.6508 ms 88.7%
  addmm 0.6877 ms 84.0%
  triton_mm_83 0.7120 ms 81.1%
  triton_mm_85 0.7370 ms 78.3%
  triton_mm_81 0.7549 ms 76.5%
  triton_mm_84 0.7806 ms 74.0%
  triton_mm_86 0.8373 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7500 seconds and 1.6147 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_108 0.5491 ms 100.0%
  triton_mm_107 0.5756 ms 95.4%
  triton_mm_109 0.5838 ms 94.1%
  bias_addmm 0.6224 ms 88.2%
  addmm 0.6360 ms 86.3%
  triton_mm_102 0.7066 ms 77.7%
  triton_mm_103 0.7390 ms 74.3%
  triton_mm_100 0.7409 ms 74.1%
  triton_mm_104 0.7531 ms 72.9%
  triton_mm_105 0.7953 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7485 seconds and 0.2147 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_122 0.0466 ms 100.0%
  triton_mm_118 0.0475 ms 98.0%
  triton_mm_126 0.0481 ms 96.7%
  triton_mm_121 0.0491 ms 94.9%
  triton_mm_114 0.0502 ms 92.8%
  triton_mm_117 0.0503 ms 92.6%
  triton_mm_111 0.0503 ms 92.5%
  bias_addmm 0.0507 ms 91.8%
  triton_mm_112 0.0513 ms 90.8%
  triton_mm_113 0.0531 ms 87.7%
SingleProcess AUTOTUNE benchmarking takes 2.2180 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_201 0.5503 ms 100.0%
  triton_mm_200 0.5767 ms 95.4%
  triton_mm_202 0.5813 ms 94.7%
  bias_addmm 0.6175 ms 89.1%
  triton_mm_195 0.7048 ms 78.1%
  addmm 0.7254 ms 75.9%
  triton_mm_196 0.7393 ms 74.4%
  triton_mm_193 0.7443 ms 73.9%
  triton_mm_197 0.7477 ms 73.6%
  triton_mm_198 0.7948 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7465 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_215 0.0466 ms 100.0%
  triton_mm_211 0.0475 ms 98.0%
  triton_mm_219 0.0487 ms 95.6%
  triton_mm_214 0.0491 ms 95.0%
  triton_mm_210 0.0502 ms 92.8%
  triton_mm_207 0.0502 ms 92.7%
  triton_mm_204 0.0503 ms 92.7%
  triton_mm_205 0.0514 ms 90.7%
  bias_addmm 0.0521 ms 89.4%
  triton_mm_206 0.0536 ms 86.9%
SingleProcess AUTOTUNE benchmarking takes 2.2218 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_308 0.0467 ms 100.0%
  triton_mm_304 0.0482 ms 96.8%
  triton_mm_312 0.0487 ms 95.9%
  triton_mm_303 0.0494 ms 94.5%
  triton_mm_297 0.0498 ms 93.8%
  triton_mm_307 0.0500 ms 93.3%
  triton_mm_300 0.0502 ms 92.9%
  triton_mm_298 0.0520 ms 89.8%
  bias_addmm 0.0522 ms 89.4%
  triton_mm_299 0.0532 ms 87.7%
SingleProcess AUTOTUNE benchmarking takes 2.2083 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_401 0.0466 ms 100.0%
  triton_mm_397 0.0475 ms 98.1%
  triton_mm_405 0.0481 ms 96.9%
  triton_mm_400 0.0490 ms 95.2%
  triton_mm_396 0.0502 ms 92.8%
  triton_mm_390 0.0504 ms 92.4%
  triton_mm_393 0.0511 ms 91.2%
  bias_addmm 0.0519 ms 89.8%
  triton_mm_391 0.0519 ms 89.8%
  triton_mm_392 0.0538 ms 86.6%
SingleProcess AUTOTUNE benchmarking takes 2.2149 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_490 0.0475 ms 100.0%
  triton_mm_494 0.0475 ms 100.0%
  triton_mm_498 0.0482 ms 98.5%
  triton_mm_493 0.0501 ms 94.8%
  triton_mm_489 0.0501 ms 94.7%
  triton_mm_483 0.0504 ms 94.2%
  triton_mm_486 0.0512 ms 92.8%
  triton_mm_484 0.0514 ms 92.5%
  bias_addmm 0.0518 ms 91.6%
  triton_mm_485 0.0531 ms 89.5%
SingleProcess AUTOTUNE benchmarking takes 2.2178 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x1920, 1920x2048)
  triton_mm_592 0.1684 ms 100.0%
  triton_mm_591 0.1688 ms 99.8%
  triton_mm_593 0.1829 ms 92.1%
  triton_mm_586 0.1901 ms 88.6%
  triton_mm_584 0.1969 ms 85.5%
  triton_mm_588 0.1998 ms 84.3%
  bias_addmm 0.2048 ms 82.2%
  triton_mm_587 0.2125 ms 79.3%
  triton_mm_589 0.2264 ms 74.4%
  triton_mm_585 0.2289 ms 73.6%
SingleProcess AUTOTUNE benchmarking takes 2.5964 seconds and 0.0283 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x2048, 2048x2048)
  triton_mm_610 0.1759 ms 100.0%
  triton_mm_611 0.1768 ms 99.5%
  bias_addmm 0.1825 ms 96.4%
  triton_mm_612 0.1908 ms 92.2%
  triton_mm_603 0.2034 ms 86.5%
  triton_mm_605 0.2035 ms 86.4%
  triton_mm_607 0.2043 ms 86.1%
  addmm 0.2096 ms 83.9%
  triton_mm_606 0.2250 ms 78.2%
  triton_mm_608 0.2387 ms 73.7%
SingleProcess AUTOTUNE benchmarking takes 2.6099 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x768)
  mm 0.0639 ms 100.0%
  triton_mm_630 0.0647 ms 98.8%
  triton_mm_629 0.0651 ms 98.1%
  triton_mm_631 0.0680 ms 94.0%
  triton_mm_625 0.0893 ms 71.6%
  triton_mm_627 0.0902 ms 70.8%
  triton_mm_623 0.0905 ms 70.6%
  triton_mm_622 0.0922 ms 69.3%
  triton_mm_626 0.0928 ms 68.9%
  triton_mm_624 0.0981 ms 65.2%
SingleProcess AUTOTUNE benchmarking takes 2.3921 seconds and 0.0012 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_644 0.0467 ms 100.0%
  triton_mm_640 0.0483 ms 96.7%
  triton_mm_648 0.0488 ms 95.6%
  triton_mm_643 0.0491 ms 95.1%
  triton_mm_633 0.0498 ms 93.8%
  triton_mm_639 0.0502 ms 92.9%
  bias_addmm 0.0508 ms 91.9%
  triton_mm_636 0.0510 ms 91.5%
  triton_mm_634 0.0522 ms 89.3%
  triton_mm_635 0.0532 ms 87.8%
SingleProcess AUTOTUNE benchmarking takes 2.2282 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(16x3840, 16x1920, 1920x3840)
  triton_mm_2872 0.0226 ms 100.0%
  triton_mm_2876 0.0226 ms 100.0%
  triton_mm_2868 0.0231 ms 97.6%
  triton_mm_2880 0.0239 ms 94.5%
  bias_addmm 0.0265 ms 85.2%
  triton_mm_2871 0.0275 ms 82.2%
  triton_mm_2875 0.0276 ms 82.0%
  addmm 0.0280 ms 80.8%
  triton_mm_2865 0.0280 ms 80.7%
  triton_mm_2867 0.0284 ms 79.4%
SingleProcess AUTOTUNE benchmarking takes 7.7908 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x32768, 4576x1920, 1920x32768)
  bias_addmm 2.4734 ms 100.0%
  triton_mm_2898 2.4912 ms 99.3%
  triton_mm_2897 2.5422 ms 97.3%
  triton_mm_2899 2.7451 ms 90.1%
  addmm 2.9545 ms 83.7%
  triton_mm_2892 3.0573 ms 80.9%
  triton_mm_2894 3.1727 ms 78.0%
  triton_mm_2890 3.2442 ms 76.2%
  triton_mm_2893 3.3171 ms 74.6%
  triton_mm_2895 3.6274 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 11.9891 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x4576, 4576x1920)
  triton_mm_2935 2.3342 ms 100.0%
  mm 2.3662 ms 98.6%
  triton_mm_2936 2.4217 ms 96.4%
  triton_mm_2937 2.5523 ms 91.5%
  triton_mm_2928 3.0643 ms 76.2%
  triton_mm_2930 3.1252 ms 74.7%
  triton_mm_2932 3.1364 ms 74.4%
  triton_mm_2931 3.2126 ms 72.7%
  triton_mm_2933 3.2443 ms 71.9%
  triton_mm_2929 3.2912 ms 70.9%
SingleProcess AUTOTUNE benchmarking takes 10.2341 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x32768, 32768x1920)
  mm 2.3293 ms 100.0%
  triton_mm_2916 2.3862 ms 97.6%
  triton_mm_2917 2.4259 ms 96.0%
  triton_mm_2918 2.4443 ms 95.3%
  triton_mm_2912 3.0953 ms 75.3%
  triton_mm_2910 3.2857 ms 70.9%
  triton_mm_2911 3.3110 ms 70.4%
  triton_mm_2909 3.3193 ms 70.2%
  triton_mm_2914 3.3987 ms 68.5%
  triton_mm_2913 3.5074 ms 66.4%
SingleProcess AUTOTUNE benchmarking takes 10.2046 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3006 0.5582 ms 100.0%
  triton_mm_3007 0.5826 ms 95.8%
  triton_mm_3008 0.6134 ms 91.0%
  mm 0.6431 ms 86.8%
  triton_mm_3003 0.6932 ms 80.5%
  triton_mm_2999 0.7121 ms 78.4%
  triton_mm_3001 0.7224 ms 77.3%
  triton_mm_3004 0.7536 ms 74.1%
  triton_mm_3002 0.7601 ms 73.4%
  triton_mm_3000 0.7690 ms 72.6%
SingleProcess AUTOTUNE benchmarking takes 9.9210 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x16, 16x1920)
  triton_mm_2960 0.0164 ms 100.0%
  triton_mm_2964 0.0164 ms 99.4%
  triton_mm_2966 0.0165 ms 99.0%
  triton_mm_2965 0.0169 ms 97.0%
  triton_mm_2961 0.0175 ms 93.5%
  triton_mm_2962 0.0176 ms 93.1%
  triton_mm_2967 0.0177 ms 92.6%
  triton_mm_2963 0.0179 ms 91.2%
  triton_mm_2959 0.0182 ms 89.8%
  triton_mm_2957 0.0182 ms 89.6%
SingleProcess AUTOTUNE benchmarking takes 5.2685 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x3840, 3840x1920)
  triton_mm_2946 0.0228 ms 100.0%
  triton_mm_2942 0.0242 ms 94.3%
  triton_mm_2950 0.0246 ms 92.6%
  mm 0.0283 ms 80.7%
  triton_mm_2954 0.0284 ms 80.5%
  triton_mm_2941 0.0342 ms 66.6%
  triton_mm_2940 0.0343 ms 66.6%
  triton_mm_2945 0.0366 ms 62.3%
  triton_mm_2949 0.0389 ms 58.6%
  triton_mm_2939 0.0406 ms 56.2%
SingleProcess AUTOTUNE benchmarking takes 2.1059 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_2987 0.5576 ms 100.0%
  mm 0.5635 ms 98.9%
  triton_mm_2988 0.5640 ms 98.9%
  triton_mm_2989 0.6206 ms 89.8%
  triton_mm_2982 0.7050 ms 79.1%
  triton_mm_2984 0.7119 ms 78.3%
  triton_mm_2980 0.7186 ms 77.6%
  triton_mm_2983 0.7659 ms 72.8%
  triton_mm_2981 0.7931 ms 70.3%
  triton_mm_2985 0.8141 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.6129 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5538 ms 100.0%
  triton_mm_3044 0.5596 ms 99.0%
  triton_mm_3045 0.5877 ms 94.2%
  triton_mm_3046 0.6164 ms 89.8%
  triton_mm_3037 0.6947 ms 79.7%
  triton_mm_3041 0.7052 ms 78.5%
  triton_mm_3039 0.7193 ms 77.0%
  triton_mm_3040 0.7533 ms 73.5%
  triton_mm_3042 0.7654 ms 72.3%
  triton_mm_3038 0.7672 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 13.0972 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3025 0.5496 ms 100.0%
  triton_mm_3026 0.5633 ms 97.6%
  triton_mm_3027 0.5754 ms 95.5%
  mm 0.5840 ms 94.1%
  triton_mm_3020 0.7025 ms 78.2%
  triton_mm_3018 0.7268 ms 75.6%
  triton_mm_3021 0.7271 ms 75.6%
  triton_mm_3022 0.7364 ms 74.6%
  triton_mm_3019 0.7519 ms 73.1%
  triton_mm_3023 0.7932 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 9.9269 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4211 ms 100.0%
  triton_mm_3120 0.4305 ms 97.8%
  triton_mm_3121 0.4523 ms 93.1%
  triton_mm_3122 0.4748 ms 88.7%
  triton_mm_3117 0.5297 ms 79.5%
  triton_mm_3113 0.5321 ms 79.1%
  triton_mm_3115 0.5425 ms 77.6%
  triton_mm_3116 0.5666 ms 74.3%
  triton_mm_3118 0.5797 ms 72.6%
  triton_mm_3114 0.5808 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 9.9482 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3102 0.4127 ms 100.0%
  triton_mm_3101 0.4131 ms 99.9%
  triton_mm_3103 0.4341 ms 95.1%
  mm 0.4422 ms 93.3%
  triton_mm_3096 0.5220 ms 79.1%
  triton_mm_3094 0.5435 ms 75.9%
  triton_mm_3097 0.5485 ms 75.2%
  triton_mm_3098 0.5502 ms 75.0%
  triton_mm_3095 0.5695 ms 72.5%
  triton_mm_3099 0.5900 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 13.1866 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1640 ms 100.0%
  triton_mm_3082 0.1860 ms 88.2%
  triton_mm_3083 0.1950 ms 84.1%
  triton_mm_3077 0.1954 ms 83.9%
  triton_mm_3075 0.2063 ms 79.5%
  triton_mm_3079 0.2064 ms 79.5%
  triton_mm_3078 0.2148 ms 76.4%
  triton_mm_3076 0.2210 ms 74.2%
  triton_mm_3080 0.2214 ms 74.1%
  triton_mm_3084 0.2336 ms 70.2%
SingleProcess AUTOTUNE benchmarking takes 9.7554 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  triton_mm_3063 0.1423 ms 100.0%
  triton_mm_3064 0.1434 ms 99.3%
  triton_mm_3065 0.1528 ms 93.2%
  mm 0.1625 ms 87.6%
  triton_mm_3058 0.1766 ms 80.6%
  triton_mm_3060 0.1849 ms 77.0%
  triton_mm_3056 0.1853 ms 76.8%
  triton_mm_3059 0.1929 ms 73.8%
  triton_mm_3057 0.1954 ms 72.8%
  triton_mm_3061 0.2024 ms 70.3%
SingleProcess AUTOTUNE benchmarking takes 10.3074 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3152 0.0328 ms 100.0%
  triton_mm_3149 0.0331 ms 99.1%
  triton_mm_3151 0.0331 ms 98.9%
  triton_mm_3147 0.0332 ms 98.8%
  triton_mm_3148 0.0336 ms 97.4%
  triton_mm_3154 0.0337 ms 97.3%
  triton_mm_3146 0.0341 ms 96.1%
  triton_mm_3145 0.0343 ms 95.4%
  triton_mm_3150 0.0348 ms 94.1%
  triton_mm_3155 0.0361 ms 90.7%
SingleProcess AUTOTUNE benchmarking takes 5.2968 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0532 ms 100.0%
  triton_mm_3131 0.0532 ms 100.0%
  triton_mm_3127 0.0535 ms 99.4%
  triton_mm_3135 0.0587 ms 90.6%
  triton_mm_3139 0.0692 ms 76.9%
  triton_mm_3126 0.0846 ms 62.8%
  triton_mm_3125 0.0885 ms 60.1%
  triton_mm_3130 0.0953 ms 55.8%
  triton_mm_3134 0.1017 ms 52.3%
  triton_mm_3124 0.1031 ms 51.6%
SingleProcess AUTOTUNE benchmarking takes 6.5896 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3172 0.5612 ms 100.0%
  mm 0.5636 ms 99.6%
  triton_mm_3173 0.5644 ms 99.4%
  triton_mm_3174 0.6145 ms 91.3%
  triton_mm_3167 0.7047 ms 79.6%
  triton_mm_3169 0.7126 ms 78.7%
  triton_mm_3165 0.7129 ms 78.7%
  triton_mm_3168 0.7660 ms 73.3%
  triton_mm_3166 0.7946 ms 70.6%
  triton_mm_3170 0.8217 ms 68.3%
SingleProcess AUTOTUNE benchmarking takes 2.6088 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3210 0.5497 ms 100.0%
  triton_mm_3211 0.5625 ms 97.7%
  triton_mm_3212 0.5760 ms 95.4%
  mm 0.5916 ms 92.9%
  triton_mm_3205 0.7138 ms 77.0%
  triton_mm_3203 0.7272 ms 75.6%
  triton_mm_3206 0.7272 ms 75.6%
  triton_mm_3207 0.7364 ms 74.6%
  triton_mm_3204 0.7675 ms 71.6%
  triton_mm_3208 0.7896 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.6279 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4190 ms 100.0%
  triton_mm_3305 0.4309 ms 97.2%
  triton_mm_3306 0.4523 ms 92.6%
  triton_mm_3307 0.4750 ms 88.2%
  triton_mm_3302 0.5319 ms 78.8%
  triton_mm_3298 0.5366 ms 78.1%
  triton_mm_3300 0.5434 ms 77.1%
  triton_mm_3301 0.5698 ms 73.5%
  triton_mm_3303 0.5797 ms 72.3%
  triton_mm_3299 0.5811 ms 72.1%
SingleProcess AUTOTUNE benchmarking takes 2.5788 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3286 0.4120 ms 100.0%
  triton_mm_3287 0.4128 ms 99.8%
  triton_mm_3288 0.4334 ms 95.0%
  mm 0.4397 ms 93.7%
  triton_mm_3281 0.5240 ms 78.6%
  triton_mm_3279 0.5459 ms 75.5%
  triton_mm_3282 0.5498 ms 74.9%
  triton_mm_3283 0.5529 ms 74.5%
  triton_mm_3280 0.5728 ms 71.9%
  triton_mm_3284 0.5941 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.5777 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1630 ms 100.0%
  triton_mm_3267 0.1848 ms 88.2%
  triton_mm_3268 0.1936 ms 84.2%
  triton_mm_3262 0.1955 ms 83.4%
  triton_mm_3269 0.1965 ms 83.0%
  triton_mm_3260 0.2059 ms 79.2%
  triton_mm_3264 0.2066 ms 78.9%
  triton_mm_3263 0.2153 ms 75.7%
  triton_mm_3261 0.2208 ms 73.8%
  triton_mm_3265 0.2213 ms 73.7%
SingleProcess AUTOTUNE benchmarking takes 2.4829 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1382 ms 100.0%
  triton_mm_3248 0.1422 ms 97.2%
  triton_mm_3249 0.1441 ms 95.9%
  triton_mm_3250 0.1528 ms 90.4%
  triton_mm_3243 0.1779 ms 77.7%
  triton_mm_3241 0.1854 ms 74.5%
  triton_mm_3245 0.1860 ms 74.3%
  triton_mm_3244 0.1923 ms 71.9%
  triton_mm_3242 0.1950 ms 70.9%
  triton_mm_3246 0.2019 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.4692 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3336 0.0331 ms 100.0%
  triton_mm_3330 0.0332 ms 99.9%
  triton_mm_3332 0.0332 ms 99.8%
  triton_mm_3337 0.0341 ms 97.2%
  triton_mm_3331 0.0341 ms 97.0%
  triton_mm_3334 0.0342 ms 96.8%
  triton_mm_3333 0.0349 ms 94.9%
  triton_mm_3335 0.0349 ms 94.9%
  triton_mm_3339 0.0349 ms 94.8%
  triton_mm_3338 0.0350 ms 94.6%
SingleProcess AUTOTUNE benchmarking takes 2.0219 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0532 ms 100.0%
  triton_mm_3312 0.0532 ms 100.0%
  triton_mm_3316 0.0532 ms 100.0%
  triton_mm_3320 0.0577 ms 92.2%
  triton_mm_3324 0.0693 ms 76.8%
  triton_mm_3311 0.0848 ms 62.7%
  triton_mm_3310 0.0886 ms 60.1%
  triton_mm_3315 0.0954 ms 55.8%
  triton_mm_3319 0.1018 ms 52.3%
  triton_mm_3309 0.1040 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 2.1603 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3357 0.5607 ms 100.0%
  mm 0.5652 ms 99.2%
  triton_mm_3358 0.5665 ms 99.0%
  triton_mm_3359 0.6191 ms 90.6%
  triton_mm_3352 0.7127 ms 78.7%
  triton_mm_3350 0.7137 ms 78.6%
  triton_mm_3354 0.7192 ms 78.0%
  triton_mm_3353 0.7662 ms 73.2%
  triton_mm_3351 0.7936 ms 70.7%
  triton_mm_3355 0.8216 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 2.6089 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3395 0.5491 ms 100.0%
  triton_mm_3396 0.5638 ms 97.4%
  triton_mm_3397 0.5753 ms 95.4%
  mm 0.5915 ms 92.8%
  triton_mm_3390 0.7022 ms 78.2%
  triton_mm_3388 0.7276 ms 75.5%
  triton_mm_3391 0.7286 ms 75.4%
  triton_mm_3392 0.7359 ms 74.6%
  triton_mm_3389 0.7602 ms 72.2%
  triton_mm_3393 0.7894 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.6266 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1380 ms 100.0%
  triton_mm_3433 0.1416 ms 97.5%
  triton_mm_3434 0.1432 ms 96.4%
  triton_mm_3435 0.1535 ms 89.9%
  triton_mm_3428 0.1771 ms 77.9%
  triton_mm_3430 0.1849 ms 74.7%
  triton_mm_3426 0.1851 ms 74.6%
  triton_mm_3429 0.1926 ms 71.7%
  triton_mm_3427 0.1951 ms 70.8%
  triton_mm_3431 0.2009 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.4842 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3580 0.5496 ms 100.0%
  triton_mm_3581 0.5639 ms 97.5%
  triton_mm_3582 0.5761 ms 95.4%
  mm 0.5913 ms 93.0%
  triton_mm_3575 0.7027 ms 78.2%
  triton_mm_3573 0.7261 ms 75.7%
  triton_mm_3576 0.7292 ms 75.4%
  triton_mm_3577 0.7358 ms 74.7%
  triton_mm_3574 0.7599 ms 72.3%
  triton_mm_3578 0.7890 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.6207 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1379 ms 100.0%
  triton_mm_3618 0.1418 ms 97.2%
  triton_mm_3619 0.1437 ms 96.0%
  triton_mm_3620 0.1524 ms 90.5%
  triton_mm_3613 0.1765 ms 78.1%
  triton_mm_3611 0.1844 ms 74.8%
  triton_mm_3615 0.1848 ms 74.6%
  triton_mm_3614 0.1918 ms 71.9%
  triton_mm_3612 0.1939 ms 71.1%
  triton_mm_3616 0.2012 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.4800 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3765 0.5499 ms 100.0%
  triton_mm_3766 0.5633 ms 97.6%
  triton_mm_3767 0.5757 ms 95.5%
  mm 0.5906 ms 93.1%
  triton_mm_3760 0.7001 ms 78.5%
  triton_mm_3758 0.7280 ms 75.5%
  triton_mm_3761 0.7288 ms 75.4%
  triton_mm_3762 0.7362 ms 74.7%
  triton_mm_3759 0.7682 ms 71.6%
  triton_mm_3763 0.7885 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.6320 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3950 0.5481 ms 100.0%
  triton_mm_3951 0.5636 ms 97.3%
  triton_mm_3952 0.5815 ms 94.3%
  mm 0.5913 ms 92.7%
  triton_mm_3945 0.7026 ms 78.0%
  triton_mm_3946 0.7241 ms 75.7%
  triton_mm_3943 0.7253 ms 75.6%
  triton_mm_3947 0.7323 ms 74.9%
  triton_mm_3944 0.7599 ms 72.1%
  triton_mm_3948 0.7891 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.6257 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_4135 0.5499 ms 100.0%
  triton_mm_4136 0.5644 ms 97.4%
  triton_mm_4137 0.5756 ms 95.5%
  mm 0.5905 ms 93.1%
  triton_mm_4130 0.7024 ms 78.3%
  triton_mm_4131 0.7244 ms 75.9%
  triton_mm_4128 0.7251 ms 75.8%
  triton_mm_4132 0.7361 ms 74.7%
  triton_mm_4129 0.7676 ms 71.6%
  triton_mm_4133 0.7891 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.6347 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x768, 768x2048)
  triton_mm_7427 0.0749 ms 100.0%
  mm 0.0756 ms 99.1%
  triton_mm_7428 0.0781 ms 95.9%
  triton_mm_7420 0.0827 ms 90.6%
  triton_mm_7424 0.0828 ms 90.4%
  triton_mm_7422 0.0829 ms 90.3%
  triton_mm_7421 0.0895 ms 83.7%
  triton_mm_7425 0.0939 ms 79.8%
  triton_mm_7429 0.0947 ms 79.1%
  triton_mm_7423 0.1082 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 10.4908 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x2048)
  triton_mm_7465 0.1715 ms 100.0%
  mm 0.1718 ms 99.9%
  triton_mm_7466 0.1802 ms 95.2%
  triton_mm_7467 0.1882 ms 91.1%
  triton_mm_7460 0.2022 ms 84.8%
  triton_mm_7462 0.2023 ms 84.8%
  triton_mm_7458 0.2024 ms 84.8%
  triton_mm_7461 0.2178 ms 78.7%
  triton_mm_7459 0.2245 ms 76.4%
  triton_mm_7463 0.2288 ms 75.0%
SingleProcess AUTOTUNE benchmarking takes 10.4493 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x2048)
  mm 0.1670 ms 100.0%
  triton_mm_7484 0.1886 ms 88.6%
  triton_mm_7485 0.1949 ms 85.7%
  triton_mm_7486 0.1987 ms 84.1%
  triton_mm_7481 0.2124 ms 78.7%
  triton_mm_7477 0.2127 ms 78.5%
  triton_mm_7482 0.2210 ms 75.6%
  triton_mm_7479 0.2212 ms 75.5%
  triton_mm_7478 0.2251 ms 74.2%
  triton_mm_7480 0.2613 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 9.8346 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x1920)
  mm 0.1648 ms 100.0%
  triton_mm_7522 0.1873 ms 88.0%
  triton_mm_7523 0.1933 ms 85.3%
  triton_mm_7524 0.1981 ms 83.2%
  triton_mm_7515 0.2065 ms 79.8%
  triton_mm_7519 0.2066 ms 79.8%
  triton_mm_7518 0.2133 ms 77.3%
  triton_mm_7517 0.2178 ms 75.7%
  triton_mm_7516 0.2218 ms 74.3%
  triton_mm_7520 0.2232 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 9.9007 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x4576, 4576x2048)
  mm 0.0684 ms 100.0%
  triton_mm_7448 0.0698 ms 98.0%
  triton_mm_7447 0.0774 ms 88.3%
  triton_mm_7441 0.0818 ms 83.6%
  triton_mm_7446 0.0865 ms 79.1%
  triton_mm_7443 0.0867 ms 78.9%
  triton_mm_7439 0.0867 ms 78.9%
  triton_mm_7442 0.0898 ms 76.2%
  triton_mm_7444 0.0900 ms 76.0%
  triton_mm_7440 0.0924 ms 74.0%
SingleProcess AUTOTUNE benchmarking takes 9.8016 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x1920)
  mm 0.1479 ms 100.0%
  triton_mm_7503 0.1513 ms 97.8%
  triton_mm_7504 0.1579 ms 93.7%
  triton_mm_7505 0.1623 ms 91.1%
  triton_mm_7498 0.1892 ms 78.2%
  triton_mm_7500 0.1952 ms 75.8%
  triton_mm_7496 0.1956 ms 75.6%
  triton_mm_7499 0.2038 ms 72.6%
  triton_mm_7497 0.2061 ms 71.8%
  triton_mm_7501 0.2124 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 10.4999 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4560, 4560x28)
  triton_mm_8646 0.0339 ms 100.0%
  triton_mm_8639 0.0345 ms 98.2%
  mm 0.0363 ms 93.4%
  triton_mm_8637 0.0431 ms 78.5%
  triton_mm_8638 0.0465 ms 72.8%
  triton_mm_8645 0.0491 ms 68.9%
  triton_mm_8636 0.0506 ms 66.9%
  triton_mm_8651 0.0507 ms 66.8%
  triton_mm_8642 0.0523 ms 64.8%
  triton_mm_8650 0.0549 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 6.7835 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(4560x28, 28x1920)
  triton_mm_8 0.0296 ms 100.0%
  triton_mm_9 0.0296 ms 100.0%
  triton_mm_7 0.0308 ms 96.3%
  triton_mm_6 0.0309 ms 95.8%
  triton_mm_11 0.0309 ms 95.7%
  triton_mm_12 0.0311 ms 95.3%
  triton_mm_4 0.0325 ms 91.1%
  triton_mm_15 0.0330 ms 89.7%
  triton_mm_16 0.0337 ms 87.9%
  triton_mm_14 0.0337 ms 87.7%
SingleProcess AUTOTUNE benchmarking takes 1.9940 seconds and 0.0081 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_29 0.0474 ms 100.0%
  triton_mm_25 0.0475 ms 99.9%
  triton_mm_33 0.0489 ms 97.1%
  triton_mm_28 0.0490 ms 96.9%
  triton_mm_24 0.0491 ms 96.7%
  triton_mm_21 0.0503 ms 94.3%
  triton_mm_18 0.0503 ms 94.2%
  triton_mm_19 0.0510 ms 93.0%
  bias_addmm 0.0517 ms 91.7%
  triton_mm_20 0.0530 ms 89.5%
SingleProcess AUTOTUNE benchmarking takes 2.2465 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_51 0.4406 ms 100.0%
  triton_mm_50 0.4465 ms 98.7%
  triton_mm_52 0.4849 ms 90.9%
  bias_addmm 0.5221 ms 84.4%
  triton_mm_45 0.5386 ms 81.8%
  triton_mm_47 0.5633 ms 78.2%
  triton_mm_43 0.5743 ms 76.7%
  triton_mm_46 0.5962 ms 73.9%
  triton_mm_48 0.6329 ms 69.6%
  triton_mm_44 0.6550 ms 67.3%
SingleProcess AUTOTUNE benchmarking takes 2.7236 seconds and 0.0152 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_70 0.1478 ms 100.0%
  triton_mm_69 0.1512 ms 97.8%
  triton_mm_71 0.1609 ms 91.9%
  bias_addmm 0.1630 ms 90.7%
  triton_mm_64 0.1801 ms 82.0%
  addmm 0.1861 ms 79.4%
  triton_mm_62 0.1881 ms 78.6%
  triton_mm_66 0.1938 ms 76.2%
  triton_mm_65 0.2017 ms 73.3%
  triton_mm_67 0.2099 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 2.5946 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(4576x7680, 4576x1920, 1920x7680)
  triton_mm_89 0.5821 ms 100.0%
  triton_mm_88 0.5902 ms 98.6%
  triton_mm_90 0.6434 ms 90.5%
  bias_addmm 0.6501 ms 89.5%
  triton_mm_83 0.7129 ms 81.7%
  triton_mm_85 0.7445 ms 78.2%
  triton_mm_81 0.7598 ms 76.6%
  addmm 0.7883 ms 73.8%
  triton_mm_84 0.7908 ms 73.6%
  triton_mm_86 0.8476 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 2.7478 seconds and 0.3334 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_108 0.5556 ms 100.0%
  triton_mm_107 0.5773 ms 96.2%
  triton_mm_109 0.5840 ms 95.1%
  bias_addmm 0.6208 ms 89.5%
  triton_mm_102 0.7063 ms 78.7%
  addmm 0.7278 ms 76.3%
  triton_mm_103 0.7340 ms 75.7%
  triton_mm_104 0.7481 ms 74.3%
  triton_mm_100 0.7495 ms 74.1%
  triton_mm_105 0.7975 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 2.7465 seconds and 0.1577 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_201 0.5508 ms 100.0%
  triton_mm_200 0.5768 ms 95.5%
  triton_mm_202 0.5824 ms 94.6%
  bias_addmm 0.6210 ms 88.7%
  triton_mm_195 0.7038 ms 78.3%
  addmm 0.7279 ms 75.7%
  triton_mm_196 0.7326 ms 75.2%
  triton_mm_193 0.7477 ms 73.7%
  triton_mm_197 0.7497 ms 73.5%
  triton_mm_198 0.7982 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7405 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_294 0.5516 ms 100.0%
  triton_mm_293 0.5774 ms 95.5%
  triton_mm_295 0.5860 ms 94.1%
  bias_addmm 0.6204 ms 88.9%
  triton_mm_288 0.7046 ms 78.3%
  addmm 0.7282 ms 75.7%
  triton_mm_289 0.7315 ms 75.4%
  triton_mm_286 0.7455 ms 74.0%
  triton_mm_290 0.7479 ms 73.8%
  triton_mm_291 0.7974 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7420 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_387 0.5505 ms 100.0%
  triton_mm_386 0.5758 ms 95.6%
  triton_mm_388 0.5812 ms 94.7%
  bias_addmm 0.6202 ms 88.8%
  addmm 0.6272 ms 87.8%
  triton_mm_381 0.7061 ms 78.0%
  triton_mm_382 0.7324 ms 75.2%
  triton_mm_379 0.7446 ms 73.9%
  triton_mm_383 0.7470 ms 73.7%
  triton_mm_384 0.7967 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.7399 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x1920, 1920x2048)
  triton_mm_592 0.1686 ms 100.0%
  triton_mm_591 0.1688 ms 99.9%
  triton_mm_593 0.1830 ms 92.1%
  triton_mm_586 0.1896 ms 88.9%
  triton_mm_584 0.1976 ms 85.3%
  triton_mm_588 0.1996 ms 84.5%
  bias_addmm 0.2044 ms 82.5%
  triton_mm_587 0.2124 ms 79.4%
  triton_mm_585 0.2280 ms 74.0%
  triton_mm_589 0.2283 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 4.6180 seconds and 0.0681 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x2048, 2048x2048)
  triton_mm_611 0.1765 ms 100.0%
  triton_mm_610 0.1772 ms 99.6%
  bias_addmm 0.1795 ms 98.4%
  triton_mm_612 0.1923 ms 91.8%
  triton_mm_603 0.2039 ms 86.6%
  triton_mm_605 0.2048 ms 86.2%
  triton_mm_607 0.2058 ms 85.8%
  addmm 0.2059 ms 85.7%
  triton_mm_606 0.2237 ms 78.9%
  triton_mm_608 0.2385 ms 74.0%
SingleProcess AUTOTUNE benchmarking takes 2.6220 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x768)
  mm 0.0639 ms 100.0%
  triton_mm_630 0.0650 ms 98.4%
  triton_mm_629 0.0656 ms 97.5%
  triton_mm_631 0.0695 ms 92.0%
  triton_mm_625 0.0890 ms 71.8%
  triton_mm_627 0.0905 ms 70.6%
  triton_mm_623 0.0908 ms 70.3%
  triton_mm_626 0.0921 ms 69.4%
  triton_mm_622 0.0927 ms 68.9%
  triton_mm_624 0.0987 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 2.4095 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(16x3840, 16x1920, 1920x3840)
  triton_mm_2872 0.0228 ms 100.0%
  triton_mm_2876 0.0230 ms 98.7%
  triton_mm_2868 0.0237 ms 96.1%
  triton_mm_2880 0.0242 ms 93.9%
  bias_addmm 0.0268 ms 84.9%
  triton_mm_2867 0.0275 ms 82.7%
  triton_mm_2871 0.0276 ms 82.6%
  triton_mm_2875 0.0278 ms 81.8%
  triton_mm_2866 0.0280 ms 81.2%
  triton_mm_2865 0.0282 ms 80.8%
SingleProcess AUTOTUNE benchmarking takes 7.6864 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x32768, 4576x1920, 1920x32768)
  triton_mm_2898 2.4253 ms 100.0%
  triton_mm_2897 2.4738 ms 98.0%
  bias_addmm 2.6001 ms 93.3%
  triton_mm_2899 2.7005 ms 89.8%
  triton_mm_2892 3.0067 ms 80.7%
  triton_mm_2894 3.0567 ms 79.3%
  triton_mm_2890 3.1635 ms 76.7%
  addmm 3.1911 ms 76.0%
  triton_mm_2893 3.2328 ms 75.0%
  triton_mm_2895 3.5005 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 12.4592 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x4576, 4576x1920)
  mm 2.2792 ms 100.0%
  triton_mm_2935 2.3200 ms 98.2%
  triton_mm_2936 2.3811 ms 95.7%
  triton_mm_2937 2.5167 ms 90.6%
  triton_mm_2928 2.9708 ms 76.7%
  triton_mm_2932 2.9884 ms 76.3%
  triton_mm_2930 3.0433 ms 74.9%
  triton_mm_2931 3.1306 ms 72.8%
  triton_mm_2933 3.2066 ms 71.1%
  triton_mm_2929 3.2532 ms 70.1%
SingleProcess AUTOTUNE benchmarking takes 13.0507 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x32768, 32768x1920)
  mm 2.2874 ms 100.0%
  triton_mm_2916 2.3550 ms 97.1%
  triton_mm_2917 2.3964 ms 95.5%
  triton_mm_2918 2.4279 ms 94.2%
  triton_mm_2912 3.0169 ms 75.8%
  triton_mm_2910 3.2627 ms 70.1%
  triton_mm_2909 3.2880 ms 69.6%
  triton_mm_2911 3.3044 ms 69.2%
  triton_mm_2914 3.3847 ms 67.6%
  triton_mm_2913 3.4468 ms 66.4%
SingleProcess AUTOTUNE benchmarking takes 10.0131 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3006 0.5523 ms 100.0%
  triton_mm_3007 0.5748 ms 96.1%
  mm 0.5790 ms 95.4%
  triton_mm_3008 0.6052 ms 91.3%
  triton_mm_3003 0.6856 ms 80.5%
  triton_mm_2999 0.6976 ms 79.2%
  triton_mm_3001 0.7141 ms 77.3%
  triton_mm_3004 0.7452 ms 74.1%
  triton_mm_3002 0.7464 ms 74.0%
  triton_mm_3000 0.7580 ms 72.9%
SingleProcess AUTOTUNE benchmarking takes 9.7915 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x16, 16x1920)
  triton_mm_2960 0.0165 ms 100.0%
  triton_mm_2962 0.0165 ms 99.8%
  triton_mm_2967 0.0167 ms 98.7%
  triton_mm_2965 0.0171 ms 96.6%
  triton_mm_2964 0.0180 ms 91.6%
  triton_mm_2966 0.0181 ms 91.2%
  triton_mm_2961 0.0182 ms 90.4%
  triton_mm_2963 0.0183 ms 89.9%
  triton_mm_2957 0.0185 ms 89.3%
  triton_mm_2968 0.0188 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 5.2943 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x3840, 3840x1920)
  triton_mm_2946 0.0233 ms 100.0%
  triton_mm_2942 0.0235 ms 99.0%
  triton_mm_2950 0.0248 ms 93.8%
  mm 0.0284 ms 81.8%
  triton_mm_2954 0.0286 ms 81.3%
  triton_mm_2941 0.0345 ms 67.4%
  triton_mm_2940 0.0353 ms 66.0%
  triton_mm_2945 0.0367 ms 63.4%
  triton_mm_2949 0.0390 ms 59.6%
  triton_mm_2939 0.0409 ms 56.9%
SingleProcess AUTOTUNE benchmarking takes 9.7433 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_2987 0.5527 ms 100.0%
  triton_mm_2988 0.5569 ms 99.3%
  mm 0.5593 ms 98.8%
  triton_mm_2989 0.6071 ms 91.0%
  triton_mm_2982 0.6891 ms 80.2%
  triton_mm_2984 0.6997 ms 79.0%
  triton_mm_2980 0.7053 ms 78.4%
  triton_mm_2983 0.7505 ms 73.7%
  triton_mm_2981 0.7845 ms 70.5%
  triton_mm_2985 0.8019 ms 68.9%
SingleProcess AUTOTUNE benchmarking takes 10.3522 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5463 ms 100.0%
  triton_mm_3044 0.5530 ms 98.8%
  triton_mm_3045 0.5798 ms 94.2%
  triton_mm_3046 0.6107 ms 89.5%
  triton_mm_3037 0.6815 ms 80.2%
  triton_mm_3041 0.6916 ms 79.0%
  triton_mm_3039 0.7071 ms 77.3%
  triton_mm_3040 0.7390 ms 73.9%
  triton_mm_3042 0.7522 ms 72.6%
  triton_mm_3038 0.7539 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 9.7690 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3025 0.5452 ms 100.0%
  triton_mm_3026 0.5577 ms 97.8%
  triton_mm_3027 0.5702 ms 95.6%
  mm 0.5860 ms 93.0%
  triton_mm_3020 0.6932 ms 78.6%
  triton_mm_3018 0.7161 ms 76.1%
  triton_mm_3021 0.7214 ms 75.6%
  triton_mm_3022 0.7264 ms 75.1%
  triton_mm_3019 0.7520 ms 72.5%
  triton_mm_3023 0.7821 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 9.8289 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4178 ms 100.0%
  triton_mm_3120 0.4262 ms 98.0%
  triton_mm_3121 0.4475 ms 93.4%
  triton_mm_3122 0.4698 ms 88.9%
  triton_mm_3117 0.5221 ms 80.0%
  triton_mm_3113 0.5256 ms 79.5%
  triton_mm_3115 0.5349 ms 78.1%
  triton_mm_3116 0.5618 ms 74.4%
  triton_mm_3118 0.5730 ms 72.9%
  triton_mm_3114 0.5739 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 9.7543 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3101 0.4090 ms 100.0%
  triton_mm_3102 0.4093 ms 99.9%
  mm 0.4360 ms 93.8%
  triton_mm_3103 0.4786 ms 85.5%
  triton_mm_3096 0.5170 ms 79.1%
  triton_mm_3094 0.5384 ms 76.0%
  triton_mm_3097 0.5399 ms 75.8%
  triton_mm_3098 0.5452 ms 75.0%
  triton_mm_3095 0.5630 ms 72.6%
  triton_mm_3099 0.5844 ms 70.0%
SingleProcess AUTOTUNE benchmarking takes 9.7422 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1622 ms 100.0%
  triton_mm_3082 0.1846 ms 87.9%
  triton_mm_3083 0.1932 ms 83.9%
  triton_mm_3077 0.1939 ms 83.7%
  triton_mm_3084 0.1945 ms 83.4%
  triton_mm_3079 0.1972 ms 82.3%
  triton_mm_3075 0.2036 ms 79.7%
  triton_mm_3078 0.2129 ms 76.2%
  triton_mm_3080 0.2201 ms 73.7%
  triton_mm_3076 0.2202 ms 73.7%
SingleProcess AUTOTUNE benchmarking takes 9.7014 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1382 ms 100.0%
  triton_mm_3063 0.1400 ms 98.7%
  triton_mm_3064 0.1418 ms 97.5%
  triton_mm_3065 0.1524 ms 90.7%
  triton_mm_3058 0.1762 ms 78.5%
  triton_mm_3056 0.1819 ms 76.0%
  triton_mm_3060 0.1839 ms 75.1%
  triton_mm_3059 0.1910 ms 72.3%
  triton_mm_3057 0.1940 ms 71.2%
  triton_mm_3061 0.2011 ms 68.7%
SingleProcess AUTOTUNE benchmarking takes 10.4689 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3152 0.0328 ms 100.0%
  triton_mm_3145 0.0333 ms 98.6%
  triton_mm_3151 0.0333 ms 98.4%
  triton_mm_3150 0.0338 ms 97.1%
  triton_mm_3149 0.0341 ms 96.2%
  triton_mm_3147 0.0345 ms 95.1%
  triton_mm_3148 0.0349 ms 93.9%
  triton_mm_3154 0.0350 ms 93.8%
  triton_mm_3155 0.0351 ms 93.4%
  triton_mm_3153 0.0352 ms 93.3%
SingleProcess AUTOTUNE benchmarking takes 8.8247 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3131 0.0526 ms 100.0%
  triton_mm_3127 0.0529 ms 99.6%
  mm 0.0531 ms 99.2%
  triton_mm_3135 0.0588 ms 89.5%
  triton_mm_3139 0.0695 ms 75.7%
  triton_mm_3126 0.0842 ms 62.5%
  triton_mm_3125 0.0886 ms 59.4%
  triton_mm_3130 0.0954 ms 55.2%
  triton_mm_3134 0.1012 ms 52.0%
  triton_mm_3124 0.1033 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 6.7475 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5467 ms 100.0%
  triton_mm_3229 0.5600 ms 97.6%
  triton_mm_3230 0.5880 ms 93.0%
  triton_mm_3231 0.6189 ms 88.3%
  triton_mm_3222 0.6945 ms 78.7%
  triton_mm_3226 0.7018 ms 77.9%
  triton_mm_3224 0.7186 ms 76.1%
  triton_mm_3225 0.7543 ms 72.5%
  triton_mm_3223 0.7653 ms 71.4%
  triton_mm_3227 0.7695 ms 71.0%
SingleProcess AUTOTUNE benchmarking takes 2.6155 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3210 0.5511 ms 100.0%
  triton_mm_3211 0.5678 ms 97.1%
  triton_mm_3212 0.5774 ms 95.4%
  mm 0.5903 ms 93.4%
  triton_mm_3205 0.7030 ms 78.4%
  triton_mm_3203 0.7283 ms 75.7%
  triton_mm_3206 0.7318 ms 75.3%
  triton_mm_3207 0.7374 ms 74.7%
  triton_mm_3204 0.7677 ms 71.8%
  triton_mm_3208 0.7924 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.6184 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3286 0.4152 ms 100.0%
  triton_mm_3287 0.4152 ms 100.0%
  triton_mm_3288 0.4354 ms 95.4%
  mm 0.4364 ms 95.1%
  triton_mm_3281 0.5252 ms 79.1%
  triton_mm_3279 0.5477 ms 75.8%
  triton_mm_3282 0.5491 ms 75.6%
  triton_mm_3283 0.5544 ms 74.9%
  triton_mm_3280 0.5705 ms 72.8%
  triton_mm_3284 0.5939 ms 69.9%
SingleProcess AUTOTUNE benchmarking takes 2.5807 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1649 ms 100.0%
  triton_mm_3267 0.1856 ms 88.8%
  triton_mm_3268 0.1941 ms 84.9%
  triton_mm_3269 0.1968 ms 83.8%
  triton_mm_3262 0.1969 ms 83.7%
  triton_mm_3260 0.2051 ms 80.4%
  triton_mm_3264 0.2057 ms 80.1%
  triton_mm_3263 0.2145 ms 76.8%
  triton_mm_3261 0.2209 ms 74.6%
  triton_mm_3265 0.2231 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 2.4881 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1389 ms 100.0%
  triton_mm_3248 0.1420 ms 97.8%
  triton_mm_3249 0.1438 ms 96.6%
  triton_mm_3250 0.1546 ms 89.8%
  triton_mm_3243 0.1775 ms 78.2%
  triton_mm_3241 0.1849 ms 75.1%
  triton_mm_3245 0.1860 ms 74.7%
  triton_mm_3244 0.1923 ms 72.2%
  triton_mm_3242 0.1953 ms 71.1%
  triton_mm_3246 0.2027 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.4800 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3337 0.0329 ms 100.0%
  triton_mm_3332 0.0334 ms 98.7%
  triton_mm_3334 0.0344 ms 95.8%
  triton_mm_3331 0.0345 ms 95.5%
  triton_mm_3330 0.0346 ms 95.1%
  triton_mm_3336 0.0347 ms 94.9%
  triton_mm_3333 0.0350 ms 94.0%
  triton_mm_3335 0.0351 ms 93.7%
  triton_mm_3339 0.0352 ms 93.5%
  triton_mm_3338 0.0365 ms 90.2%
SingleProcess AUTOTUNE benchmarking takes 2.0409 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3312 0.0527 ms 100.0%
  triton_mm_3316 0.0532 ms 99.2%
  mm 0.0532 ms 99.0%
  triton_mm_3320 0.0575 ms 91.7%
  triton_mm_3324 0.0694 ms 76.0%
  triton_mm_3311 0.0842 ms 62.6%
  triton_mm_3310 0.0879 ms 60.0%
  triton_mm_3315 0.0956 ms 55.2%
  triton_mm_3319 0.1020 ms 51.7%
  triton_mm_3309 0.1034 ms 51.0%
SingleProcess AUTOTUNE benchmarking takes 2.1624 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5474 ms 100.0%
  triton_mm_3414 0.5612 ms 97.5%
  triton_mm_3415 0.5907 ms 92.7%
  triton_mm_3416 0.6195 ms 88.4%
  triton_mm_3407 0.6966 ms 78.6%
  triton_mm_3411 0.7087 ms 77.2%
  triton_mm_3409 0.7175 ms 76.3%
  triton_mm_3410 0.7553 ms 72.5%
  triton_mm_3412 0.7631 ms 71.7%
  triton_mm_3408 0.7718 ms 70.9%
SingleProcess AUTOTUNE benchmarking takes 2.6155 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3395 0.5516 ms 100.0%
  triton_mm_3396 0.5673 ms 97.2%
  triton_mm_3397 0.5784 ms 95.4%
  mm 0.5914 ms 93.3%
  triton_mm_3390 0.7079 ms 77.9%
  triton_mm_3388 0.7289 ms 75.7%
  triton_mm_3391 0.7344 ms 75.1%
  triton_mm_3392 0.7404 ms 74.5%
  triton_mm_3389 0.7675 ms 71.9%
  triton_mm_3393 0.7924 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.6306 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1645 ms 100.0%
  triton_mm_3452 0.1859 ms 88.5%
  triton_mm_3453 0.1944 ms 84.6%
  triton_mm_3454 0.1972 ms 83.4%
  triton_mm_3447 0.1973 ms 83.3%
  triton_mm_3449 0.2045 ms 80.4%
  triton_mm_3445 0.2070 ms 79.5%
  triton_mm_3448 0.2147 ms 76.6%
  triton_mm_3446 0.2215 ms 74.2%
  triton_mm_3450 0.2226 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 2.4873 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1381 ms 100.0%
  triton_mm_3433 0.1433 ms 96.4%
  triton_mm_3434 0.1447 ms 95.4%
  triton_mm_3435 0.1534 ms 90.1%
  triton_mm_3428 0.1785 ms 77.4%
  triton_mm_3426 0.1841 ms 75.0%
  triton_mm_3430 0.1861 ms 74.2%
  triton_mm_3429 0.1929 ms 71.6%
  triton_mm_3427 0.1962 ms 70.4%
  triton_mm_3431 0.2035 ms 67.9%
SingleProcess AUTOTUNE benchmarking takes 2.4934 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3515 0.0335 ms 100.0%
  triton_mm_3521 0.0335 ms 99.9%
  triton_mm_3517 0.0336 ms 99.7%
  triton_mm_3522 0.0345 ms 97.0%
  triton_mm_3519 0.0345 ms 96.9%
  triton_mm_3518 0.0353 ms 94.9%
  triton_mm_3520 0.0353 ms 94.9%
  triton_mm_3524 0.0353 ms 94.7%
  triton_mm_3516 0.0359 ms 93.1%
  triton_mm_3523 0.0367 ms 91.3%
SingleProcess AUTOTUNE benchmarking takes 2.0412 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3497 0.0528 ms 100.0%
  triton_mm_3501 0.0532 ms 99.2%
  mm 0.0532 ms 99.2%
  triton_mm_3505 0.0575 ms 91.9%
  triton_mm_3509 0.0696 ms 75.9%
  triton_mm_3496 0.0842 ms 62.7%
  triton_mm_3495 0.0887 ms 59.5%
  triton_mm_3500 0.0956 ms 55.3%
  triton_mm_3504 0.1021 ms 51.7%
  triton_mm_3494 0.1034 ms 51.1%
SingleProcess AUTOTUNE benchmarking takes 2.1717 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5470 ms 100.0%
  triton_mm_3599 0.5608 ms 97.5%
  triton_mm_3600 0.5899 ms 92.7%
  triton_mm_3601 0.6192 ms 88.3%
  triton_mm_3592 0.6940 ms 78.8%
  triton_mm_3596 0.7019 ms 77.9%
  triton_mm_3594 0.7177 ms 76.2%
  triton_mm_3595 0.7494 ms 73.0%
  triton_mm_3593 0.7650 ms 71.5%
  triton_mm_3597 0.7694 ms 71.1%
SingleProcess AUTOTUNE benchmarking takes 2.6128 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3580 0.5513 ms 100.0%
  triton_mm_3581 0.5658 ms 97.4%
  triton_mm_3582 0.5778 ms 95.4%
  mm 0.5904 ms 93.4%
  triton_mm_3575 0.7030 ms 78.4%
  triton_mm_3573 0.7285 ms 75.7%
  triton_mm_3576 0.7316 ms 75.3%
  triton_mm_3577 0.7380 ms 74.7%
  triton_mm_3574 0.7680 ms 71.8%
  triton_mm_3578 0.7915 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 2.6199 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5512 ms 100.0%
  triton_mm_3784 0.5588 ms 98.6%
  triton_mm_3785 0.5873 ms 93.9%
  triton_mm_3786 0.6183 ms 89.1%
  triton_mm_3777 0.6991 ms 78.8%
  triton_mm_3781 0.7003 ms 78.7%
  triton_mm_3779 0.7156 ms 77.0%
  triton_mm_3780 0.7545 ms 73.1%
  triton_mm_3782 0.7617 ms 72.4%
  triton_mm_3778 0.7712 ms 71.5%
SingleProcess AUTOTUNE benchmarking takes 2.6196 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3765 0.5509 ms 100.0%
  triton_mm_3766 0.5656 ms 97.4%
  triton_mm_3767 0.5775 ms 95.4%
  mm 0.5899 ms 93.4%
  triton_mm_3760 0.7027 ms 78.4%
  triton_mm_3758 0.7303 ms 75.4%
  triton_mm_3761 0.7343 ms 75.0%
  triton_mm_3762 0.7397 ms 74.5%
  triton_mm_3759 0.7708 ms 71.5%
  triton_mm_3763 0.7945 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 2.6291 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x768, 768x2048)
  triton_mm_7427 0.0746 ms 100.0%
  triton_mm_7428 0.0779 ms 95.7%
  triton_mm_7424 0.0828 ms 90.1%
  triton_mm_7420 0.0844 ms 88.4%
  triton_mm_7422 0.0852 ms 87.5%
  triton_mm_7421 0.0920 ms 81.0%
  mm 0.0934 ms 79.8%
  triton_mm_7425 0.0943 ms 79.0%
  triton_mm_7429 0.0974 ms 76.6%
  triton_mm_7423 0.1110 ms 67.2%
SingleProcess AUTOTUNE benchmarking takes 10.6369 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x2048)
  mm 0.1707 ms 100.0%
  triton_mm_7465 0.1726 ms 98.9%
  triton_mm_7466 0.1796 ms 95.1%
  triton_mm_7467 0.1874 ms 91.1%
  triton_mm_7462 0.1990 ms 85.8%
  triton_mm_7458 0.2009 ms 85.0%
  triton_mm_7460 0.2016 ms 84.7%
  triton_mm_7461 0.2193 ms 77.8%
  triton_mm_7459 0.2243 ms 76.1%
  triton_mm_7463 0.2292 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 10.4856 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x2048)
  mm 0.1653 ms 100.0%
  triton_mm_7484 0.1886 ms 87.7%
  triton_mm_7485 0.1934 ms 85.5%
  triton_mm_7481 0.2084 ms 79.3%
  triton_mm_7477 0.2100 ms 78.7%
  triton_mm_7480 0.2187 ms 75.6%
  triton_mm_7479 0.2200 ms 75.2%
  triton_mm_7482 0.2218 ms 74.6%
  triton_mm_7478 0.2247 ms 73.6%
  triton_mm_7486 0.2374 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 9.9561 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x1920)
  mm 0.1658 ms 100.0%
  triton_mm_7522 0.1868 ms 88.8%
  triton_mm_7523 0.1944 ms 85.3%
  triton_mm_7524 0.1985 ms 83.6%
  triton_mm_7519 0.2051 ms 80.9%
  triton_mm_7515 0.2055 ms 80.7%
  triton_mm_7518 0.2147 ms 77.2%
  triton_mm_7517 0.2168 ms 76.5%
  triton_mm_7516 0.2228 ms 74.5%
  triton_mm_7520 0.2230 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 9.9290 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x4576, 4576x2048)
  mm 0.0683 ms 100.0%
  triton_mm_7448 0.0706 ms 96.6%
  triton_mm_7447 0.0788 ms 86.7%
  triton_mm_7441 0.0815 ms 83.8%
  triton_mm_7446 0.0869 ms 78.5%
  triton_mm_7439 0.0875 ms 78.0%
  triton_mm_7443 0.0890 ms 76.7%
  triton_mm_7442 0.0912 ms 74.9%
  triton_mm_7444 0.0914 ms 74.7%
  triton_mm_7440 0.0941 ms 72.5%
SingleProcess AUTOTUNE benchmarking takes 9.8508 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x1920)
  mm 0.1458 ms 100.0%
  triton_mm_7503 0.1514 ms 96.3%
  triton_mm_7504 0.1580 ms 92.3%
  triton_mm_7505 0.1628 ms 89.6%
  triton_mm_7498 0.1905 ms 76.5%
  triton_mm_7496 0.1960 ms 74.4%
  triton_mm_7500 0.1963 ms 74.3%
  triton_mm_7499 0.2072 ms 70.4%
  triton_mm_7497 0.2082 ms 70.0%
  triton_mm_7501 0.2142 ms 68.1%
SingleProcess AUTOTUNE benchmarking takes 2.4980 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(1920x4560, 4560x28)
  triton_mm_8639 0.0349 ms 100.0%
  triton_mm_8646 0.0351 ms 99.3%
  mm 0.0416 ms 83.8%
  triton_mm_8637 0.0441 ms 79.1%
  triton_mm_8638 0.0486 ms 71.7%
  triton_mm_8645 0.0513 ms 68.0%
  triton_mm_8651 0.0516 ms 67.5%
  triton_mm_8636 0.0527 ms 66.2%
  triton_mm_8642 0.0536 ms 65.0%
  triton_mm_8644 0.0567 ms 61.5%
SingleProcess AUTOTUNE benchmarking takes 2.1057 seconds and 0.0815 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(4560x28, 28x1920)
  triton_mm_7 0.0292 ms 100.0%
  triton_mm_12 0.0297 ms 98.3%
  triton_mm_8 0.0304 ms 96.1%
  triton_mm_9 0.0304 ms 96.0%
  triton_mm_6 0.0304 ms 95.9%
  triton_mm_11 0.0305 ms 95.8%
  triton_mm_15 0.0319 ms 91.6%
  triton_mm_3 0.0319 ms 91.5%
  triton_mm_4 0.0319 ms 91.5%
  triton_mm_16 0.0328 ms 88.9%
SingleProcess AUTOTUNE benchmarking takes 2.0459 seconds and 0.0113 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_29 0.0465 ms 100.0%
  triton_mm_25 0.0474 ms 98.1%
  triton_mm_33 0.0478 ms 97.3%
  triton_mm_28 0.0489 ms 95.2%
  triton_mm_24 0.0499 ms 93.3%
  triton_mm_18 0.0502 ms 92.8%
  bias_addmm 0.0506 ms 91.9%
  triton_mm_21 0.0509 ms 91.4%
  triton_mm_19 0.0511 ms 91.0%
  triton_mm_20 0.0531 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.2048 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_51 0.4342 ms 100.0%
  triton_mm_50 0.4428 ms 98.0%
  triton_mm_52 0.4786 ms 90.7%
  bias_addmm 0.5214 ms 83.3%
  triton_mm_45 0.5359 ms 81.0%
  triton_mm_47 0.5514 ms 78.7%
  triton_mm_43 0.5658 ms 76.7%
  triton_mm_46 0.5911 ms 73.5%
  addmm 0.5988 ms 72.5%
  triton_mm_48 0.6270 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7162 seconds and 0.0271 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_70 0.1462 ms 100.0%
  triton_mm_69 0.1499 ms 97.5%
  triton_mm_71 0.1588 ms 92.1%
  bias_addmm 0.1612 ms 90.7%
  triton_mm_64 0.1786 ms 81.9%
  addmm 0.1843 ms 79.3%
  triton_mm_62 0.1861 ms 78.6%
  triton_mm_66 0.1896 ms 77.1%
  triton_mm_65 0.1996 ms 73.2%
  triton_mm_67 0.2082 ms 70.2%
SingleProcess AUTOTUNE benchmarking takes 2.5904 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(4576x7680, 4576x1920, 1920x7680)
  triton_mm_89 0.5790 ms 100.0%
  triton_mm_88 0.5885 ms 98.4%
  triton_mm_90 0.6400 ms 90.5%
  bias_addmm 0.6509 ms 89.0%
  triton_mm_83 0.7127 ms 81.2%
  triton_mm_85 0.7339 ms 78.9%
  triton_mm_81 0.7566 ms 76.5%
  triton_mm_84 0.7863 ms 73.6%
  addmm 0.7880 ms 73.5%
  triton_mm_86 0.8377 ms 69.1%
SingleProcess AUTOTUNE benchmarking takes 2.7490 seconds and 0.3361 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_108 0.5511 ms 100.0%
  triton_mm_107 0.5754 ms 95.8%
  triton_mm_109 0.5829 ms 94.5%
  bias_addmm 0.6197 ms 88.9%
  triton_mm_102 0.7045 ms 78.2%
  addmm 0.7278 ms 75.7%
  triton_mm_103 0.7369 ms 74.8%
  triton_mm_100 0.7398 ms 74.5%
  triton_mm_104 0.7483 ms 73.7%
  triton_mm_105 0.7985 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 2.7501 seconds and 0.1785 seconds precompiling
AUTOTUNE addmm(16x11520, 16x1920, 1920x11520)
  triton_mm_122 0.0465 ms 100.0%
  triton_mm_126 0.0479 ms 97.1%
  triton_mm_118 0.0481 ms 96.7%
  triton_mm_117 0.0492 ms 94.5%
  triton_mm_121 0.0498 ms 93.3%
  triton_mm_111 0.0499 ms 93.1%
  triton_mm_114 0.0510 ms 91.2%
  triton_mm_112 0.0512 ms 90.8%
  bias_addmm 0.0515 ms 90.3%
  triton_mm_113 0.0531 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.2281 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x5760, 4576x1920, 1920x5760)
  triton_mm_144 0.4351 ms 100.0%
  triton_mm_143 0.4425 ms 98.3%
  triton_mm_145 0.4798 ms 90.7%
  bias_addmm 0.5197 ms 83.7%
  triton_mm_138 0.5339 ms 81.5%
  triton_mm_140 0.5549 ms 78.4%
  triton_mm_136 0.5685 ms 76.5%
  triton_mm_139 0.5896 ms 73.8%
  triton_mm_141 0.6305 ms 69.0%
  triton_mm_137 0.6424 ms 67.7%
SingleProcess AUTOTUNE benchmarking takes 2.7072 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_163 0.1459 ms 100.0%
  triton_mm_162 0.1498 ms 97.4%
  triton_mm_164 0.1584 ms 92.1%
  bias_addmm 0.1612 ms 90.5%
  triton_mm_157 0.1777 ms 82.1%
  addmm 0.1842 ms 79.2%
  triton_mm_155 0.1852 ms 78.8%
  triton_mm_159 0.1895 ms 77.0%
  triton_mm_158 0.2002 ms 72.9%
  triton_mm_160 0.2078 ms 70.2%
SingleProcess AUTOTUNE benchmarking takes 2.5874 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x7680, 7680x1920)
  triton_mm_201 0.5520 ms 100.0%
  triton_mm_200 0.5780 ms 95.5%
  triton_mm_202 0.5872 ms 94.0%
  bias_addmm 0.6206 ms 89.0%
  addmm 0.6326 ms 87.3%
  triton_mm_195 0.7028 ms 78.5%
  triton_mm_196 0.7383 ms 74.8%
  triton_mm_193 0.7426 ms 74.3%
  triton_mm_197 0.7479 ms 73.8%
  triton_mm_198 0.7972 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 2.7509 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_256 0.1472 ms 100.0%
  triton_mm_255 0.1508 ms 97.6%
  triton_mm_257 0.1598 ms 92.1%
  triton_mm_250 0.1778 ms 82.8%
  addmm 0.1846 ms 79.8%
  triton_mm_248 0.1853 ms 79.5%
  bias_addmm 0.1885 ms 78.1%
  triton_mm_252 0.1903 ms 77.4%
  triton_mm_251 0.1998 ms 73.7%
  triton_mm_253 0.2093 ms 70.4%
SingleProcess AUTOTUNE benchmarking takes 2.5780 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_349 0.1464 ms 100.0%
  triton_mm_348 0.1504 ms 97.3%
  triton_mm_350 0.1590 ms 92.1%
  triton_mm_343 0.1772 ms 82.6%
  triton_mm_341 0.1855 ms 78.9%
  bias_addmm 0.1885 ms 77.6%
  triton_mm_345 0.1909 ms 76.7%
  triton_mm_344 0.2003 ms 73.1%
  triton_mm_346 0.2090 ms 70.0%
  triton_mm_342 0.2105 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.5926 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_442 0.1462 ms 100.0%
  triton_mm_441 0.1504 ms 97.2%
  triton_mm_443 0.1587 ms 92.1%
  triton_mm_436 0.1777 ms 82.3%
  triton_mm_434 0.1853 ms 78.9%
  bias_addmm 0.1886 ms 77.5%
  triton_mm_438 0.1900 ms 77.0%
  triton_mm_437 0.2002 ms 73.0%
  triton_mm_439 0.2080 ms 70.3%
  triton_mm_435 0.2105 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.5844 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x1920, 4576x1920, 1920x1920)
  triton_mm_535 0.1462 ms 100.0%
  triton_mm_534 0.1500 ms 97.5%
  triton_mm_536 0.1587 ms 92.1%
  triton_mm_529 0.1791 ms 81.6%
  triton_mm_527 0.1852 ms 79.0%
  bias_addmm 0.1877 ms 77.9%
  triton_mm_531 0.1901 ms 76.9%
  triton_mm_530 0.2000 ms 73.1%
  triton_mm_532 0.2086 ms 70.1%
  triton_mm_528 0.2105 ms 69.5%
SingleProcess AUTOTUNE benchmarking takes 2.6076 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x1920, 1920x2048)
  triton_mm_592 0.1689 ms 100.0%
  triton_mm_591 0.1696 ms 99.6%
  triton_mm_593 0.1837 ms 91.9%
  triton_mm_586 0.1893 ms 89.2%
  triton_mm_584 0.1963 ms 86.0%
  triton_mm_588 0.1988 ms 85.0%
  bias_addmm 0.2031 ms 83.2%
  triton_mm_587 0.2110 ms 80.0%
  triton_mm_589 0.2274 ms 74.3%
  triton_mm_585 0.2288 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 2.5973 seconds and 0.0177 seconds precompiling
AUTOTUNE addmm(4576x2048, 4576x2048, 2048x2048)
  triton_mm_610 0.1780 ms 100.0%
  bias_addmm 0.1824 ms 97.6%
  triton_mm_612 0.1922 ms 92.6%
  triton_mm_605 0.2044 ms 87.1%
  triton_mm_603 0.2055 ms 86.6%
  addmm 0.2095 ms 85.0%
  triton_mm_611 0.2095 ms 85.0%
  triton_mm_606 0.2241 ms 79.4%
  triton_mm_608 0.2392 ms 74.4%
  triton_mm_607 0.2412 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 10.9328 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x768)
  mm 0.0629 ms 100.0%
  triton_mm_630 0.0637 ms 98.7%
  triton_mm_629 0.0643 ms 97.9%
  triton_mm_631 0.0681 ms 92.3%
  triton_mm_625 0.0888 ms 70.8%
  triton_mm_627 0.0905 ms 69.5%
  triton_mm_623 0.0907 ms 69.4%
  triton_mm_626 0.0915 ms 68.8%
  triton_mm_622 0.0920 ms 68.4%
  triton_mm_624 0.0979 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.4019 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(16x3840, 16x1920, 1920x3840)
  triton_mm_2876 0.0226 ms 100.0%
  triton_mm_2872 0.0226 ms 99.9%
  triton_mm_2880 0.0228 ms 99.2%
  triton_mm_2868 0.0235 ms 96.2%
  bias_addmm 0.0251 ms 90.0%
  triton_mm_2871 0.0262 ms 86.1%
  addmm 0.0274 ms 82.5%
  triton_mm_2875 0.0274 ms 82.3%
  triton_mm_2865 0.0278 ms 81.2%
  triton_mm_2867 0.0284 ms 79.6%
SingleProcess AUTOTUNE benchmarking takes 8.0311 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(4576x32768, 4576x1920, 1920x32768)
  triton_mm_2898 2.4887 ms 100.0%
  triton_mm_2897 2.5087 ms 99.2%
  bias_addmm 2.6006 ms 95.7%
  triton_mm_2899 2.7428 ms 90.7%
  addmm 2.9504 ms 84.4%
  triton_mm_2892 3.0501 ms 81.6%
  triton_mm_2894 3.1274 ms 79.6%
  triton_mm_2890 3.2097 ms 77.5%
  triton_mm_2893 3.3058 ms 75.3%
  triton_mm_2895 3.5724 ms 69.7%
SingleProcess AUTOTUNE benchmarking takes 12.2511 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(32768x4576, 4576x1920)
  mm 2.3129 ms 100.0%
  triton_mm_2935 2.3297 ms 99.3%
  triton_mm_2936 2.4015 ms 96.3%
  triton_mm_2937 2.5492 ms 90.7%
  triton_mm_2928 2.9915 ms 77.3%
  triton_mm_2932 3.0867 ms 74.9%
  triton_mm_2930 3.1135 ms 74.3%
  triton_mm_2931 3.2043 ms 72.2%
  triton_mm_2933 3.2516 ms 71.1%
  triton_mm_2929 3.2944 ms 70.2%
SingleProcess AUTOTUNE benchmarking takes 9.9849 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x32768, 32768x1920)
  mm 2.3144 ms 100.0%
  triton_mm_2916 2.3635 ms 97.9%
  triton_mm_2917 2.4256 ms 95.4%
  triton_mm_2918 2.4399 ms 94.9%
  triton_mm_2912 3.0776 ms 75.2%
  triton_mm_2909 3.2927 ms 70.3%
  triton_mm_2910 3.3041 ms 70.0%
  triton_mm_2911 3.3089 ms 69.9%
  triton_mm_2914 3.3915 ms 68.2%
  triton_mm_2913 3.4960 ms 66.2%
SingleProcess AUTOTUNE benchmarking takes 10.3680 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x7680)
  triton_mm_3006 0.5580 ms 100.0%
  mm 0.5805 ms 96.1%
  triton_mm_3007 0.5840 ms 95.6%
  triton_mm_3008 0.6135 ms 91.0%
  triton_mm_3003 0.6892 ms 81.0%
  triton_mm_2999 0.7057 ms 79.1%
  triton_mm_3001 0.7199 ms 77.5%
  triton_mm_3004 0.7531 ms 74.1%
  triton_mm_3002 0.7541 ms 74.0%
  triton_mm_3000 0.7676 ms 72.7%
SingleProcess AUTOTUNE benchmarking takes 9.8752 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(3840x16, 16x1920)
  triton_mm_2960 0.0162 ms 100.0%
  triton_mm_2962 0.0162 ms 100.0%
  triton_mm_2967 0.0165 ms 98.4%
  triton_mm_2966 0.0172 ms 94.4%
  triton_mm_2964 0.0174 ms 93.4%
  triton_mm_2961 0.0174 ms 93.3%
  triton_mm_2963 0.0176 ms 92.3%
  triton_mm_2965 0.0176 ms 92.0%
  triton_mm_2957 0.0182 ms 89.1%
  triton_mm_2970 0.0186 ms 87.4%
SingleProcess AUTOTUNE benchmarking takes 5.4038 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x3840, 3840x1920)
  triton_mm_2946 0.0227 ms 100.0%
  triton_mm_2942 0.0245 ms 92.6%
  triton_mm_2950 0.0245 ms 92.4%
  mm 0.0281 ms 80.7%
  triton_mm_2954 0.0284 ms 79.8%
  triton_mm_2940 0.0349 ms 65.0%
  triton_mm_2941 0.0350 ms 64.7%
  triton_mm_2945 0.0365 ms 62.1%
  triton_mm_2949 0.0389 ms 58.4%
  triton_mm_2939 0.0407 ms 55.7%
SingleProcess AUTOTUNE benchmarking takes 6.5794 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_2987 0.5585 ms 100.0%
  triton_mm_2988 0.5639 ms 99.1%
  mm 0.5658 ms 98.7%
  triton_mm_2989 0.6166 ms 90.6%
  triton_mm_2982 0.6920 ms 80.7%
  triton_mm_2984 0.7069 ms 79.0%
  triton_mm_2980 0.7146 ms 78.2%
  triton_mm_2983 0.7625 ms 73.3%
  triton_mm_2981 0.7851 ms 71.1%
  triton_mm_2985 0.8138 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 10.2916 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5526 ms 100.0%
  triton_mm_3044 0.5609 ms 98.5%
  triton_mm_3045 0.5878 ms 94.0%
  triton_mm_3046 0.6192 ms 89.2%
  triton_mm_3037 0.6938 ms 79.6%
  triton_mm_3041 0.7018 ms 78.7%
  triton_mm_3039 0.7153 ms 77.2%
  triton_mm_3040 0.7483 ms 73.8%
  triton_mm_3042 0.7617 ms 72.5%
  triton_mm_3038 0.7677 ms 72.0%
SingleProcess AUTOTUNE benchmarking takes 9.7791 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x7680, 7680x1920)
  triton_mm_3025 0.5509 ms 100.0%
  triton_mm_3026 0.5635 ms 97.8%
  triton_mm_3027 0.5765 ms 95.6%
  mm 0.5924 ms 93.0%
  triton_mm_3020 0.7012 ms 78.6%
  triton_mm_3018 0.7249 ms 76.0%
  triton_mm_3021 0.7272 ms 75.8%
  triton_mm_3022 0.7325 ms 75.2%
  triton_mm_3019 0.7559 ms 72.9%
  triton_mm_3023 0.7916 ms 69.6%
SingleProcess AUTOTUNE benchmarking takes 9.9758 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4228 ms 100.0%
  triton_mm_3120 0.4303 ms 98.3%
  triton_mm_3121 0.4538 ms 93.2%
  triton_mm_3122 0.4756 ms 88.9%
  triton_mm_3117 0.5284 ms 80.0%
  triton_mm_3113 0.5336 ms 79.2%
  triton_mm_3115 0.5407 ms 78.2%
  triton_mm_3116 0.5689 ms 74.3%
  triton_mm_3114 0.5746 ms 73.6%
  triton_mm_3118 0.5805 ms 72.8%
SingleProcess AUTOTUNE benchmarking takes 9.8258 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3102 0.4131 ms 100.0%
  triton_mm_3101 0.4132 ms 100.0%
  triton_mm_3103 0.4335 ms 95.3%
  mm 0.4396 ms 94.0%
  triton_mm_3096 0.5211 ms 79.3%
  triton_mm_3094 0.5421 ms 76.2%
  triton_mm_3097 0.5476 ms 75.4%
  triton_mm_3098 0.5520 ms 74.8%
  triton_mm_3095 0.5688 ms 72.6%
  triton_mm_3099 0.5923 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 9.9611 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1632 ms 100.0%
  triton_mm_3082 0.1853 ms 88.1%
  triton_mm_3083 0.1945 ms 83.9%
  triton_mm_3077 0.1960 ms 83.3%
  triton_mm_3084 0.1960 ms 83.3%
  triton_mm_3079 0.2049 ms 79.7%
  triton_mm_3075 0.2052 ms 79.6%
  triton_mm_3078 0.2158 ms 75.6%
  triton_mm_3076 0.2212 ms 73.8%
  triton_mm_3080 0.2227 ms 73.3%
SingleProcess AUTOTUNE benchmarking takes 9.8399 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1396 ms 100.0%
  triton_mm_3063 0.1417 ms 98.6%
  triton_mm_3064 0.1441 ms 96.9%
  triton_mm_3065 0.1538 ms 90.8%
  triton_mm_3058 0.1776 ms 78.6%
  triton_mm_3056 0.1828 ms 76.4%
  triton_mm_3060 0.1844 ms 75.7%
  triton_mm_3059 0.1926 ms 72.5%
  triton_mm_3057 0.1948 ms 71.7%
  triton_mm_3061 0.2023 ms 69.0%
SingleProcess AUTOTUNE benchmarking takes 10.2964 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3152 0.0323 ms 100.0%
  triton_mm_3151 0.0327 ms 98.8%
  triton_mm_3149 0.0328 ms 98.5%
  triton_mm_3147 0.0329 ms 98.3%
  triton_mm_3146 0.0337 ms 95.8%
  triton_mm_3145 0.0338 ms 95.7%
  triton_mm_3148 0.0345 ms 93.6%
  triton_mm_3150 0.0345 ms 93.6%
  triton_mm_3154 0.0347 ms 93.3%
  triton_mm_3155 0.0349 ms 92.5%
SingleProcess AUTOTUNE benchmarking takes 8.9948 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0528 ms 100.0%
  triton_mm_3127 0.0534 ms 98.9%
  triton_mm_3131 0.0534 ms 98.9%
  triton_mm_3135 0.0585 ms 90.3%
  triton_mm_3139 0.0691 ms 76.5%
  triton_mm_3126 0.0847 ms 62.4%
  triton_mm_3125 0.0880 ms 60.0%
  triton_mm_3130 0.0949 ms 55.7%
  triton_mm_3134 0.1012 ms 52.2%
  triton_mm_3124 0.1040 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 6.5262 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x7680)
  triton_mm_3172 0.5608 ms 100.0%
  mm 0.5659 ms 99.1%
  triton_mm_3173 0.5679 ms 98.7%
  triton_mm_3174 0.6200 ms 90.5%
  triton_mm_3167 0.7012 ms 80.0%
  triton_mm_3169 0.7123 ms 78.7%
  triton_mm_3165 0.7141 ms 78.5%
  triton_mm_3168 0.7724 ms 72.6%
  triton_mm_3166 0.8015 ms 70.0%
  triton_mm_3170 0.8148 ms 68.8%
SingleProcess AUTOTUNE benchmarking takes 2.6156 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5539 ms 100.0%
  triton_mm_3229 0.5596 ms 99.0%
  triton_mm_3230 0.5894 ms 94.0%
  triton_mm_3231 0.6217 ms 89.1%
  triton_mm_3222 0.6960 ms 79.6%
  triton_mm_3226 0.7026 ms 78.8%
  triton_mm_3224 0.7174 ms 77.2%
  triton_mm_3225 0.7496 ms 73.9%
  triton_mm_3227 0.7643 ms 72.5%
  triton_mm_3223 0.7737 ms 71.6%
SingleProcess AUTOTUNE benchmarking takes 2.6173 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4199 ms 100.0%
  triton_mm_3305 0.4310 ms 97.4%
  triton_mm_3306 0.4548 ms 92.3%
  triton_mm_3307 0.4770 ms 88.0%
  triton_mm_3302 0.5315 ms 79.0%
  triton_mm_3298 0.5366 ms 78.3%
  triton_mm_3300 0.5422 ms 77.4%
  triton_mm_3301 0.5707 ms 73.6%
  triton_mm_3303 0.5809 ms 72.3%
  triton_mm_3299 0.5830 ms 72.0%
SingleProcess AUTOTUNE benchmarking takes 2.5828 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3287 0.4143 ms 100.0%
  triton_mm_3286 0.4153 ms 99.8%
  triton_mm_3288 0.4360 ms 95.0%
  mm 0.4419 ms 93.8%
  triton_mm_3281 0.5232 ms 79.2%
  triton_mm_3279 0.5462 ms 75.8%
  triton_mm_3282 0.5503 ms 75.3%
  triton_mm_3283 0.5540 ms 74.8%
  triton_mm_3280 0.5721 ms 72.4%
  triton_mm_3284 0.5936 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 2.5860 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1632 ms 100.0%
  triton_mm_3267 0.1857 ms 87.9%
  triton_mm_3268 0.1942 ms 84.0%
  triton_mm_3269 0.1964 ms 83.1%
  triton_mm_3262 0.1966 ms 83.0%
  triton_mm_3264 0.2048 ms 79.7%
  triton_mm_3260 0.2055 ms 79.4%
  triton_mm_3263 0.2163 ms 75.4%
  triton_mm_3261 0.2211 ms 73.8%
  triton_mm_3265 0.2229 ms 73.2%
SingleProcess AUTOTUNE benchmarking takes 2.4831 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1388 ms 100.0%
  triton_mm_3248 0.1420 ms 97.8%
  triton_mm_3249 0.1438 ms 96.6%
  triton_mm_3250 0.1545 ms 89.9%
  triton_mm_3243 0.1773 ms 78.3%
  triton_mm_3241 0.1836 ms 75.6%
  triton_mm_3245 0.1850 ms 75.1%
  triton_mm_3244 0.1918 ms 72.4%
  triton_mm_3242 0.1953 ms 71.1%
  triton_mm_3246 0.2026 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 2.4853 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(11520x16, 16x1920)
  triton_mm_3337 0.0323 ms 100.0%
  triton_mm_3336 0.0327 ms 98.8%
  triton_mm_3332 0.0328 ms 98.6%
  triton_mm_3334 0.0329 ms 98.2%
  triton_mm_3330 0.0338 ms 95.6%
  triton_mm_3331 0.0338 ms 95.6%
  triton_mm_3333 0.0345 ms 93.7%
  triton_mm_3335 0.0345 ms 93.6%
  triton_mm_3339 0.0347 ms 93.2%
  triton_mm_3340 0.0350 ms 92.4%
SingleProcess AUTOTUNE benchmarking takes 2.0330 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0528 ms 100.0%
  triton_mm_3312 0.0531 ms 99.5%
  triton_mm_3316 0.0531 ms 99.5%
  triton_mm_3320 0.0575 ms 91.9%
  triton_mm_3324 0.0681 ms 77.6%
  triton_mm_3311 0.0840 ms 62.9%
  triton_mm_3310 0.0881 ms 60.0%
  triton_mm_3315 0.0949 ms 55.7%
  triton_mm_3319 0.1018 ms 51.9%
  triton_mm_3309 0.1033 ms 51.1%
SingleProcess AUTOTUNE benchmarking takes 2.1619 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5512 ms 100.0%
  triton_mm_3414 0.5610 ms 98.2%
  triton_mm_3415 0.5890 ms 93.6%
  triton_mm_3416 0.6201 ms 88.9%
  triton_mm_3407 0.6968 ms 79.1%
  triton_mm_3411 0.7064 ms 78.0%
  triton_mm_3409 0.7173 ms 76.8%
  triton_mm_3410 0.7533 ms 73.2%
  triton_mm_3412 0.7696 ms 71.6%
  triton_mm_3408 0.7729 ms 71.3%
SingleProcess AUTOTUNE benchmarking takes 2.6205 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4208 ms 100.0%
  triton_mm_3490 0.4305 ms 97.8%
  triton_mm_3491 0.4542 ms 92.6%
  triton_mm_3492 0.4784 ms 88.0%
  triton_mm_3487 0.5302 ms 79.4%
  triton_mm_3483 0.5358 ms 78.5%
  triton_mm_3485 0.5432 ms 77.5%
  triton_mm_3486 0.5713 ms 73.7%
  triton_mm_3484 0.5818 ms 72.3%
  triton_mm_3488 0.5827 ms 72.2%
SingleProcess AUTOTUNE benchmarking takes 2.5905 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3471 0.4147 ms 100.0%
  triton_mm_3472 0.4161 ms 99.7%
  triton_mm_3473 0.4367 ms 94.9%
  mm 0.4411 ms 94.0%
  triton_mm_3466 0.5240 ms 79.1%
  triton_mm_3464 0.5447 ms 76.1%
  triton_mm_3467 0.5497 ms 75.4%
  triton_mm_3468 0.5550 ms 74.7%
  triton_mm_3465 0.5727 ms 72.4%
  triton_mm_3469 0.5944 ms 69.8%
SingleProcess AUTOTUNE benchmarking takes 2.5867 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1641 ms 100.0%
  triton_mm_3452 0.1857 ms 88.4%
  triton_mm_3453 0.1950 ms 84.2%
  triton_mm_3447 0.1952 ms 84.1%
  triton_mm_3454 0.1957 ms 83.9%
  triton_mm_3449 0.2041 ms 80.4%
  triton_mm_3445 0.2060 ms 79.7%
  triton_mm_3448 0.2156 ms 76.1%
  triton_mm_3446 0.2215 ms 74.1%
  triton_mm_3450 0.2220 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 2.4850 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1389 ms 100.0%
  triton_mm_3433 0.1425 ms 97.5%
  triton_mm_3434 0.1442 ms 96.4%
  triton_mm_3435 0.1534 ms 90.6%
  triton_mm_3428 0.1775 ms 78.2%
  triton_mm_3426 0.1842 ms 75.4%
  triton_mm_3430 0.1849 ms 75.1%
  triton_mm_3429 0.1924 ms 72.2%
  triton_mm_3427 0.1952 ms 71.2%
  triton_mm_3431 0.2024 ms 68.6%
SingleProcess AUTOTUNE benchmarking takes 2.4815 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  triton_mm_3497 0.0529 ms 100.0%
  mm 0.0530 ms 99.9%
  triton_mm_3501 0.0535 ms 99.0%
  triton_mm_3505 0.0587 ms 90.1%
  triton_mm_3509 0.0680 ms 77.8%
  triton_mm_3496 0.0848 ms 62.4%
  triton_mm_3495 0.0887 ms 59.7%
  triton_mm_3500 0.0949 ms 55.8%
  triton_mm_3504 0.1019 ms 51.9%
  triton_mm_3494 0.1033 ms 51.2%
SingleProcess AUTOTUNE benchmarking takes 2.1660 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(7680x4576, 4576x1920)
  mm 0.5497 ms 100.0%
  triton_mm_3599 0.5610 ms 98.0%
  triton_mm_3600 0.5892 ms 93.3%
  triton_mm_3601 0.6205 ms 88.6%
  triton_mm_3592 0.6914 ms 79.5%
  triton_mm_3596 0.7011 ms 78.4%
  triton_mm_3594 0.7161 ms 76.8%
  triton_mm_3595 0.7541 ms 72.9%
  triton_mm_3597 0.7625 ms 72.1%
  triton_mm_3593 0.7713 ms 71.3%
SingleProcess AUTOTUNE benchmarking takes 2.6265 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5760x4576, 4576x1920)
  mm 0.4240 ms 100.0%
  triton_mm_3675 0.4309 ms 98.4%
  triton_mm_3676 0.4544 ms 93.3%
  triton_mm_3677 0.4779 ms 88.7%
  triton_mm_3672 0.5304 ms 79.9%
  triton_mm_3668 0.5345 ms 79.3%
  triton_mm_3670 0.5406 ms 78.4%
  triton_mm_3671 0.5709 ms 74.3%
  triton_mm_3673 0.5816 ms 72.9%
  triton_mm_3669 0.5828 ms 72.7%
SingleProcess AUTOTUNE benchmarking takes 2.5851 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x5760, 5760x1920)
  triton_mm_3656 0.4141 ms 100.0%
  triton_mm_3657 0.4150 ms 99.8%
  triton_mm_3658 0.4365 ms 94.9%
  mm 0.4406 ms 94.0%
  triton_mm_3651 0.5243 ms 79.0%
  triton_mm_3649 0.5451 ms 76.0%
  triton_mm_3652 0.5498 ms 75.3%
  triton_mm_3653 0.5545 ms 74.7%
  triton_mm_3650 0.5742 ms 72.1%
  triton_mm_3654 0.5968 ms 69.4%
SingleProcess AUTOTUNE benchmarking takes 2.5872 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4576, 4576x1920)
  mm 0.1632 ms 100.0%
  triton_mm_3637 0.1850 ms 88.2%
  triton_mm_3638 0.1942 ms 84.0%
  triton_mm_3632 0.1956 ms 83.5%
  triton_mm_3639 0.1962 ms 83.2%
  triton_mm_3634 0.1981 ms 82.4%
  triton_mm_3630 0.2047 ms 79.8%
  triton_mm_3633 0.2155 ms 75.7%
  triton_mm_3631 0.2213 ms 73.8%
  triton_mm_3635 0.2220 ms 73.5%
SingleProcess AUTOTUNE benchmarking takes 2.4871 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x1920, 1920x1920)
  mm 0.1386 ms 100.0%
  triton_mm_3618 0.1419 ms 97.7%
  triton_mm_3619 0.1436 ms 96.5%
  triton_mm_3620 0.1544 ms 89.8%
  triton_mm_3613 0.1773 ms 78.1%
  triton_mm_3611 0.1838 ms 75.4%
  triton_mm_3615 0.1852 ms 74.8%
  triton_mm_3614 0.1924 ms 72.0%
  triton_mm_3612 0.1951 ms 71.0%
  triton_mm_3616 0.2026 ms 68.4%
SingleProcess AUTOTUNE benchmarking takes 2.4792 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(16x11520, 11520x1920)
  mm 0.0528 ms 100.0%
  triton_mm_3686 0.0531 ms 99.6%
  triton_mm_3682 0.0532 ms 99.2%
  triton_mm_3690 0.0586 ms 90.1%
  triton_mm_3694 0.0690 ms 76.5%
  triton_mm_3681 0.0847 ms 62.4%
  triton_mm_3680 0.0880 ms 60.0%
  triton_mm_3685 0.0949 ms 55.7%
  triton_mm_3689 0.1012 ms 52.2%
  triton_mm_3679 0.1040 ms 50.8%
SingleProcess AUTOTUNE benchmarking takes 2.1638 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x768, 768x2048)
  triton_mm_7427 0.0749 ms 100.0%
  mm 0.0758 ms 98.8%
  triton_mm_7424 0.0817 ms 91.7%
  triton_mm_7420 0.0823 ms 91.0%
  triton_mm_7422 0.0833 ms 89.9%
  triton_mm_7421 0.0893 ms 83.8%
  triton_mm_7428 0.0928 ms 80.7%
  triton_mm_7425 0.0938 ms 79.8%
  triton_mm_7423 0.1088 ms 68.8%
  triton_mm_7417 0.1130 ms 66.3%
SingleProcess AUTOTUNE benchmarking takes 10.7124 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x2048)
  mm 0.1696 ms 100.0%
  triton_mm_7465 0.1712 ms 99.1%
  triton_mm_7466 0.1797 ms 94.4%
  triton_mm_7467 0.1879 ms 90.3%
  triton_mm_7462 0.2004 ms 84.7%
  triton_mm_7458 0.2007 ms 84.5%
  triton_mm_7460 0.2025 ms 83.8%
  triton_mm_7461 0.2178 ms 77.9%
  triton_mm_7459 0.2257 ms 75.2%
  triton_mm_7463 0.2276 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 10.3693 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x2048)
  mm 0.1677 ms 100.0%
  triton_mm_7484 0.1879 ms 89.3%
  triton_mm_7485 0.1934 ms 86.7%
  triton_mm_7486 0.1983 ms 84.6%
  triton_mm_7481 0.2091 ms 80.2%
  triton_mm_7477 0.2108 ms 79.6%
  triton_mm_7480 0.2172 ms 77.2%
  triton_mm_7479 0.2186 ms 76.7%
  triton_mm_7482 0.2206 ms 76.0%
  triton_mm_7478 0.2244 ms 74.8%
SingleProcess AUTOTUNE benchmarking takes 9.8319 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(2048x4576, 4576x1920)
  mm 0.1645 ms 100.0%
  triton_mm_7522 0.1867 ms 88.1%
  triton_mm_7523 0.1935 ms 85.0%
  triton_mm_7524 0.1979 ms 83.1%
  triton_mm_7519 0.2046 ms 80.4%
  triton_mm_7515 0.2048 ms 80.3%
  triton_mm_7518 0.2134 ms 77.1%
  triton_mm_7517 0.2169 ms 75.8%
  triton_mm_7516 0.2220 ms 74.1%
  triton_mm_7520 0.2228 ms 73.8%
SingleProcess AUTOTUNE benchmarking takes 9.8789 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(768x4576, 4576x2048)
  mm 0.0686 ms 100.0%
  triton_mm_7448 0.0697 ms 98.4%
  triton_mm_7447 0.0764 ms 89.8%
  triton_mm_7441 0.0818 ms 83.9%
  triton_mm_7443 0.0862 ms 79.6%
  triton_mm_7446 0.0864 ms 79.4%
  triton_mm_7439 0.0867 ms 79.2%
  triton_mm_7442 0.0887 ms 77.4%
  triton_mm_7444 0.0906 ms 75.7%
  triton_mm_7440 0.0911 ms 75.3%
SingleProcess AUTOTUNE benchmarking takes 9.7054 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(4576x2048, 2048x1920)
  mm 0.1469 ms 100.0%
  triton_mm_7503 0.1498 ms 98.1%
  triton_mm_7504 0.1571 ms 93.5%
  triton_mm_7505 0.1608 ms 91.3%
  triton_mm_7498 0.1889 ms 77.7%
  triton_mm_7500 0.1942 ms 75.6%
  triton_mm_7496 0.1947 ms 75.4%
  triton_mm_7499 0.2032 ms 72.3%
  triton_mm_7497 0.2056 ms 71.5%
  triton_mm_7501 0.2118 ms 69.3%
SingleProcess AUTOTUNE benchmarking takes 10.4270 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(1920x4560, 4560x28)
  triton_mm_8639 0.0336 ms 100.0%
  triton_mm_8646 0.0338 ms 99.5%
  mm 0.0361 ms 93.3%
  triton_mm_8637 0.0422 ms 79.7%
  triton_mm_8638 0.0474 ms 71.0%
  triton_mm_8645 0.0492 ms 68.3%
  triton_mm_8651 0.0506 ms 66.5%
  triton_mm_8642 0.0515 ms 65.4%
  triton_mm_8636 0.0516 ms 65.2%
  triton_mm_8650 0.0556 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 7.1330 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_8757 1.6701 ms 100.0%
  mm 1.8090 ms 92.3%
  triton_mm_8754 2.0346 ms 82.1%
  triton_mm_8758 2.0672 ms 80.8%
  triton_mm_8750 2.0808 ms 80.3%
  triton_mm_8755 2.2054 ms 75.7%
  triton_mm_8752 2.2669 ms 73.7%
  triton_mm_8751 2.2774 ms 73.3%
  triton_mm_8747 2.7496 ms 60.7%
  triton_mm_8748 2.8765 ms 58.1%
SingleProcess AUTOTUNE benchmarking takes 2.2863 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6614 ms 100.0%
  triton_mm_8850 1.6803 ms 98.9%
  triton_mm_8847 2.0361 ms 81.6%
  triton_mm_8851 2.0995 ms 79.1%
  triton_mm_8843 2.1056 ms 78.9%
  triton_mm_8848 2.2293 ms 74.5%
  triton_mm_8844 2.2512 ms 73.8%
  triton_mm_8845 2.2727 ms 73.1%
  triton_mm_8840 2.7910 ms 59.5%
  triton_mm_8841 2.8771 ms 57.7%
SingleProcess AUTOTUNE benchmarking takes 2.2988 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6683 ms 100.0%
  triton_mm_8943 1.6725 ms 99.7%
  triton_mm_8940 2.0350 ms 82.0%
  triton_mm_8944 2.0859 ms 80.0%
  triton_mm_8936 2.1035 ms 79.3%
  triton_mm_8941 2.2289 ms 74.8%
  triton_mm_8938 2.2684 ms 73.5%
  triton_mm_8937 2.2722 ms 73.4%
  triton_mm_8933 2.7791 ms 60.0%
  triton_mm_8934 2.8759 ms 58.0%
SingleProcess AUTOTUNE benchmarking takes 2.3025 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6664 ms 100.0%
  triton_mm_9036 1.6732 ms 99.6%
  triton_mm_9033 2.0364 ms 81.8%
  triton_mm_9037 2.0867 ms 79.9%
  triton_mm_9029 2.1048 ms 79.2%
  triton_mm_9034 2.2286 ms 74.8%
  triton_mm_9031 2.2714 ms 73.4%
  triton_mm_9030 2.2765 ms 73.2%
  triton_mm_9026 2.7793 ms 60.0%
  triton_mm_9027 2.8776 ms 57.9%
SingleProcess AUTOTUNE benchmarking takes 2.3065 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6554 ms 100.0%
  triton_mm_9501 1.6872 ms 98.1%
  triton_mm_9498 2.0363 ms 81.3%
  triton_mm_9502 2.0863 ms 79.3%
  triton_mm_9494 2.1048 ms 78.6%
  triton_mm_9499 2.2284 ms 74.3%
  triton_mm_9495 2.2513 ms 73.5%
  triton_mm_9496 2.2684 ms 73.0%
  triton_mm_9491 2.7796 ms 59.6%
  triton_mm_9492 2.8760 ms 57.6%
SingleProcess AUTOTUNE benchmarking takes 2.2990 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6620 ms 100.0%
  triton_mm_10803 1.6706 ms 99.5%
  triton_mm_10800 2.0364 ms 81.6%
  triton_mm_10804 2.0919 ms 79.5%
  triton_mm_10796 2.1048 ms 79.0%
  triton_mm_10801 2.2294 ms 74.5%
  triton_mm_10797 2.2511 ms 73.8%
  triton_mm_10798 2.2701 ms 73.2%
  triton_mm_10793 2.7907 ms 59.6%
  triton_mm_10794 2.8900 ms 57.5%
SingleProcess AUTOTUNE benchmarking takes 2.2981 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6646 ms 100.0%
  triton_mm_10896 1.6712 ms 99.6%
  triton_mm_10893 2.0351 ms 81.8%
  triton_mm_10897 2.0792 ms 80.1%
  triton_mm_10889 2.1044 ms 79.1%
  triton_mm_10894 2.2283 ms 74.7%
  triton_mm_10890 2.2506 ms 74.0%
  triton_mm_10891 2.2707 ms 73.3%
  triton_mm_10886 2.7797 ms 59.9%
  triton_mm_10887 2.8763 ms 57.9%
SingleProcess AUTOTUNE benchmarking takes 2.2982 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6630 ms 100.0%
  triton_mm_11361 1.6698 ms 99.6%
  triton_mm_11358 2.0365 ms 81.7%
  triton_mm_11362 2.0867 ms 79.7%
  triton_mm_11354 2.1046 ms 79.0%
  triton_mm_11359 2.2298 ms 74.6%
  triton_mm_11355 2.2510 ms 73.9%
  triton_mm_11356 2.2919 ms 72.6%
  triton_mm_11351 2.7800 ms 59.8%
  triton_mm_11360 2.8982 ms 57.4%
SingleProcess AUTOTUNE benchmarking takes 2.3057 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6641 ms 100.0%
  triton_mm_11454 1.6711 ms 99.6%
  triton_mm_11451 2.0350 ms 81.8%
  triton_mm_11455 2.0923 ms 79.5%
  triton_mm_11447 2.1148 ms 78.7%
  triton_mm_11452 2.2290 ms 74.7%
  triton_mm_11448 2.2588 ms 73.7%
  triton_mm_11449 2.3005 ms 72.3%
  triton_mm_11444 2.7913 ms 59.6%
  triton_mm_11445 2.8890 ms 57.6%
SingleProcess AUTOTUNE benchmarking takes 2.3016 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x2048)
  mm 0.4421 ms 100.0%
  triton_mm_11511 0.4523 ms 97.7%
  triton_mm_11512 0.5568 ms 79.4%
  triton_mm_11508 0.5570 ms 79.4%
  triton_mm_11504 0.5591 ms 79.1%
  triton_mm_11509 0.6009 ms 73.6%
  triton_mm_11505 0.6014 ms 73.5%
  triton_mm_11506 0.6135 ms 72.1%
  triton_mm_11501 0.7503 ms 58.9%
  triton_mm_11502 0.7741 ms 57.1%
SingleProcess AUTOTUNE benchmarking takes 2.1006 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x2048, 2048x2048)
  mm 0.4637 ms 100.0%
  triton_mm_11530 0.4774 ms 97.1%
  triton_mm_11531 0.5902 ms 78.6%
  triton_mm_11527 0.5916 ms 78.4%
  triton_mm_11523 0.5964 ms 77.8%
  triton_mm_11524 0.6308 ms 73.5%
  triton_mm_11528 0.6319 ms 73.4%
  triton_mm_11525 0.6453 ms 71.9%
  triton_mm_11520 0.7947 ms 58.4%
  triton_mm_11521 0.8212 ms 56.5%
SingleProcess AUTOTUNE benchmarking takes 2.0989 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_8663 0.0742 ms 100.0%
  triton_mm_8659 0.0755 ms 98.3%
  triton_mm_8668 0.0761 ms 97.6%
  triton_mm_8662 0.0771 ms 96.2%
  triton_mm_8666 0.0774 ms 95.9%
  triton_mm_8654 0.0775 ms 95.7%
  triton_mm_8665 0.0780 ms 95.2%
  triton_mm_8661 0.0783 ms 94.8%
  mm 0.0785 ms 94.5%
  triton_mm_8653 0.0785 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.9161 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6840x28, 28x1920)
  triton_mm_8676 0.0390 ms 100.0%
  triton_mm_8677 0.0395 ms 98.8%
  triton_mm_8675 0.0395 ms 98.7%
  triton_mm_8678 0.0397 ms 98.3%
  triton_mm_8681 0.0399 ms 97.7%
  triton_mm_8680 0.0407 ms 95.8%
  triton_mm_8673 0.0412 ms 94.6%
  triton_mm_8684 0.0433 ms 90.2%
  triton_mm_8672 0.0434 ms 89.8%
  triton_mm_8671 0.0452 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 2.1505 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8702 1.2668 ms 100.0%
  bias_addmm 1.2971 ms 97.7%
  addmm 1.4649 ms 86.5%
  triton_mm_8699 1.5345 ms 82.6%
  triton_mm_8695 1.5709 ms 80.6%
  triton_mm_8703 1.5722 ms 80.6%
  triton_mm_8700 1.6749 ms 75.6%
  triton_mm_8696 1.6959 ms 74.7%
  triton_mm_8697 1.7121 ms 74.0%
  triton_mm_8692 2.0920 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.3657 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8721 0.4413 ms 100.0%
  mm 0.4556 ms 96.9%
  triton_mm_8714 0.5202 ms 84.8%
  triton_mm_8718 0.5263 ms 83.9%
  triton_mm_8722 0.5372 ms 82.2%
  triton_mm_8715 0.5592 ms 78.9%
  triton_mm_8719 0.5625 ms 78.5%
  triton_mm_8716 0.5738 ms 76.9%
  triton_mm_8711 0.7003 ms 63.0%
  triton_mm_8712 0.7226 ms 61.1%
SingleProcess AUTOTUNE benchmarking takes 2.0949 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8776 1.6953 ms 100.0%
  mm 1.7168 ms 98.7%
  triton_mm_8777 1.9887 ms 85.2%
  triton_mm_8769 2.1293 ms 79.6%
  triton_mm_8773 2.1536 ms 78.7%
  triton_mm_8774 2.1870 ms 77.5%
  triton_mm_8770 2.2094 ms 76.7%
  triton_mm_8771 2.2232 ms 76.3%
  triton_mm_8766 2.7980 ms 60.6%
  triton_mm_8767 2.8477 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.3100 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9744 0.4412 ms 100.0%
  mm 0.4540 ms 97.2%
  triton_mm_9737 0.5199 ms 84.9%
  triton_mm_9741 0.5256 ms 83.9%
  triton_mm_9745 0.5391 ms 81.8%
  triton_mm_9738 0.5596 ms 78.8%
  triton_mm_9742 0.5629 ms 78.4%
  triton_mm_9739 0.5743 ms 76.8%
  triton_mm_9734 0.7005 ms 63.0%
  triton_mm_9735 0.7265 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0925 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9799 1.6930 ms 100.0%
  mm 1.7171 ms 98.6%
  triton_mm_9800 1.9866 ms 85.2%
  triton_mm_9792 2.1292 ms 79.5%
  triton_mm_9796 2.1528 ms 78.6%
  triton_mm_9793 2.2085 ms 76.7%
  triton_mm_9797 2.2088 ms 76.6%
  triton_mm_9794 2.2274 ms 76.0%
  triton_mm_9789 2.7978 ms 60.5%
  triton_mm_9790 2.8448 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.2871 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_10116 0.4409 ms 100.0%
  mm 0.4542 ms 97.1%
  triton_mm_10109 0.5209 ms 84.6%
  triton_mm_10113 0.5243 ms 84.1%
  triton_mm_10117 0.5387 ms 81.8%
  triton_mm_10110 0.5609 ms 78.6%
  triton_mm_10114 0.5629 ms 78.3%
  triton_mm_10111 0.5747 ms 76.7%
  triton_mm_10106 0.7012 ms 62.9%
  triton_mm_10107 0.7263 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0879 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_10130 0.0743 ms 100.0%
  triton_mm_10126 0.0762 ms 97.6%
  triton_mm_10135 0.0762 ms 97.5%
  triton_mm_10129 0.0763 ms 97.5%
  triton_mm_10121 0.0765 ms 97.1%
  triton_mm_10133 0.0773 ms 96.1%
  triton_mm_10128 0.0775 ms 95.9%
  triton_mm_10120 0.0782 ms 95.1%
  triton_mm_10132 0.0784 ms 94.8%
  mm 0.0786 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.9263 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x3840)
  triton_mm_11428 0.0390 ms 100.0%
  triton_mm_11424 0.0390 ms 99.9%
  triton_mm_11423 0.0395 ms 98.6%
  triton_mm_11437 0.0401 ms 97.3%
  triton_mm_11422 0.0402 ms 97.1%
  triton_mm_11425 0.0406 ms 95.9%
  triton_mm_11435 0.0408 ms 95.6%
  triton_mm_11432 0.0411 ms 94.9%
  triton_mm_11431 0.0432 ms 90.2%
  mm 0.0442 ms 88.3%
SingleProcess AUTOTUNE benchmarking takes 1.9057 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(6864x32768, 6864x1920, 1920x32768)
  triton_mm_11492 7.0398 ms 100.0%
  bias_addmm 7.0784 ms 99.5%
  addmm 7.9854 ms 88.2%
  triton_mm_11489 8.7312 ms 80.6%
  triton_mm_11493 8.7981 ms 80.0%
  triton_mm_11485 8.9385 ms 78.8%
  triton_mm_11490 9.4261 ms 74.7%
  triton_mm_11487 9.6178 ms 73.2%
  triton_mm_11486 9.6194 ms 73.2%
  triton_mm_11482 11.8284 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 3.2799 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(6864x768, 6864x2048, 2048x768)
  triton_mm_11549 0.1839 ms 100.0%
  bias_addmm 0.1864 ms 98.7%
  addmm 0.2125 ms 86.6%
  triton_mm_11550 0.2210 ms 83.2%
  triton_mm_11542 0.2249 ms 81.8%
  triton_mm_11546 0.2265 ms 81.2%
  triton_mm_11543 0.2367 ms 77.7%
  triton_mm_11547 0.2396 ms 76.8%
  triton_mm_11544 0.2437 ms 75.5%
  triton_mm_11539 0.3038 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.1592 seconds and 0.0010 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:39<1:54:36, 159.93s/it]AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_8757 1.6492 ms 100.0%
  mm 1.8184 ms 90.7%
  triton_mm_8754 2.0170 ms 81.8%
  triton_mm_8758 2.0453 ms 80.6%
  triton_mm_8750 2.0850 ms 79.1%
  triton_mm_8755 2.1778 ms 75.7%
  triton_mm_8751 2.2264 ms 74.1%
  triton_mm_8752 2.2340 ms 73.8%
  triton_mm_8747 2.7599 ms 59.8%
  triton_mm_8748 2.8518 ms 57.8%
SingleProcess AUTOTUNE benchmarking takes 2.2723 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6474 ms 100.0%
  triton_mm_8850 1.6490 ms 99.9%
  triton_mm_8847 2.0262 ms 81.3%
  triton_mm_8851 2.0415 ms 80.7%
  triton_mm_8843 2.0833 ms 79.1%
  triton_mm_8848 2.2107 ms 74.5%
  triton_mm_8844 2.2263 ms 74.0%
  triton_mm_8845 2.2619 ms 72.8%
  triton_mm_8840 2.7282 ms 60.4%
  triton_mm_8841 2.8190 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 2.2802 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6242 ms 100.0%
  triton_mm_9222 1.6491 ms 98.5%
  triton_mm_9219 2.0184 ms 80.5%
  triton_mm_9223 2.0550 ms 79.0%
  triton_mm_9215 2.1093 ms 77.0%
  triton_mm_9220 2.1782 ms 74.6%
  triton_mm_9216 2.2271 ms 72.9%
  triton_mm_9217 2.2529 ms 72.1%
  triton_mm_9212 2.7321 ms 59.4%
  triton_mm_9213 2.8226 ms 57.5%
SingleProcess AUTOTUNE benchmarking takes 2.2717 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6472 ms 100.0%
  triton_mm_11175 1.6498 ms 99.8%
  triton_mm_11172 1.9960 ms 82.5%
  triton_mm_11176 2.0423 ms 80.7%
  triton_mm_11168 2.0855 ms 79.0%
  triton_mm_11173 2.2025 ms 74.8%
  triton_mm_11169 2.2286 ms 73.9%
  triton_mm_11170 2.2337 ms 73.7%
  triton_mm_11165 2.7314 ms 60.3%
  triton_mm_11166 2.8225 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 2.2767 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x2048)
  mm 0.4309 ms 100.0%
  triton_mm_11511 0.4422 ms 97.4%
  triton_mm_11512 0.5465 ms 78.8%
  triton_mm_11504 0.5508 ms 78.2%
  triton_mm_11508 0.5510 ms 78.2%
  triton_mm_11509 0.5896 ms 73.1%
  triton_mm_11505 0.5937 ms 72.6%
  triton_mm_11506 0.6052 ms 71.2%
  triton_mm_11501 0.7393 ms 58.3%
  triton_mm_11502 0.7599 ms 56.7%
SingleProcess AUTOTUNE benchmarking takes 2.0882 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x2048, 2048x2048)
  mm 0.4596 ms 100.0%
  triton_mm_11530 0.4676 ms 98.3%
  triton_mm_11531 0.5743 ms 80.0%
  triton_mm_11527 0.5822 ms 78.9%
  triton_mm_11523 0.5883 ms 78.1%
  triton_mm_11524 0.6182 ms 74.3%
  triton_mm_11528 0.6210 ms 74.0%
  triton_mm_11525 0.6348 ms 72.4%
  triton_mm_11520 0.7853 ms 58.5%
  triton_mm_11521 0.8116 ms 56.6%
SingleProcess AUTOTUNE benchmarking takes 2.0868 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_8663 0.0743 ms 100.0%
  triton_mm_8659 0.0753 ms 98.6%
  triton_mm_8668 0.0758 ms 98.1%
  triton_mm_8662 0.0764 ms 97.3%
  triton_mm_8666 0.0775 ms 95.9%
  triton_mm_8661 0.0776 ms 95.7%
  mm 0.0780 ms 95.2%
  triton_mm_8654 0.0781 ms 95.1%
  triton_mm_8665 0.0788 ms 94.4%
  triton_mm_8656 0.0797 ms 93.2%
SingleProcess AUTOTUNE benchmarking takes 1.9030 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(6840x28, 28x1920)
  triton_mm_8678 0.0390 ms 100.0%
  triton_mm_8680 0.0402 ms 97.0%
  triton_mm_8677 0.0404 ms 96.7%
  triton_mm_8675 0.0404 ms 96.6%
  triton_mm_8676 0.0405 ms 96.4%
  triton_mm_8681 0.0410 ms 95.2%
  triton_mm_8673 0.0428 ms 91.2%
  triton_mm_8684 0.0430 ms 90.6%
  triton_mm_8671 0.0441 ms 88.4%
  triton_mm_8672 0.0449 ms 86.9%
SingleProcess AUTOTUNE benchmarking takes 2.1275 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8702 1.2490 ms 100.0%
  bias_addmm 1.2804 ms 97.5%
  addmm 1.4472 ms 86.3%
  triton_mm_8699 1.5208 ms 82.1%
  triton_mm_8695 1.5544 ms 80.4%
  triton_mm_8703 1.5636 ms 79.9%
  triton_mm_8700 1.6547 ms 75.5%
  triton_mm_8696 1.6764 ms 74.5%
  triton_mm_8697 1.7009 ms 73.4%
  triton_mm_8692 2.0782 ms 60.1%
SingleProcess AUTOTUNE benchmarking takes 2.3604 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8721 0.4326 ms 100.0%
  mm 0.4459 ms 97.0%
  triton_mm_8714 0.5131 ms 84.3%
  triton_mm_8718 0.5153 ms 84.0%
  triton_mm_8722 0.5315 ms 81.4%
  triton_mm_8715 0.5513 ms 78.5%
  triton_mm_8719 0.5550 ms 78.0%
  triton_mm_8716 0.5640 ms 76.7%
  triton_mm_8711 0.6933 ms 62.4%
  triton_mm_8712 0.7181 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 2.0799 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8776 1.6661 ms 100.0%
  mm 1.6885 ms 98.7%
  triton_mm_8777 1.9588 ms 85.1%
  triton_mm_8773 2.1297 ms 78.2%
  triton_mm_8769 2.1323 ms 78.1%
  triton_mm_8774 2.1995 ms 75.8%
  triton_mm_8770 2.2110 ms 75.4%
  triton_mm_8771 2.2114 ms 75.3%
  triton_mm_8766 2.7740 ms 60.1%
  triton_mm_8767 2.8175 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2773 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9186 0.4340 ms 100.0%
  mm 0.4442 ms 97.7%
  triton_mm_9179 0.5144 ms 84.4%
  triton_mm_9183 0.5187 ms 83.7%
  triton_mm_9187 0.5326 ms 81.5%
  triton_mm_9184 0.5549 ms 78.2%
  triton_mm_9180 0.5556 ms 78.1%
  triton_mm_9181 0.5663 ms 76.6%
  triton_mm_9176 0.6949 ms 62.4%
  triton_mm_9177 0.7189 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.0806 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9241 1.6841 ms 100.0%
  mm 1.6895 ms 99.7%
  triton_mm_9242 1.9645 ms 85.7%
  triton_mm_9234 2.1089 ms 79.9%
  triton_mm_9238 2.1252 ms 79.2%
  triton_mm_9239 2.1872 ms 77.0%
  triton_mm_9236 2.2043 ms 76.4%
  triton_mm_9235 2.2124 ms 76.1%
  triton_mm_9231 2.8004 ms 60.1%
  triton_mm_9232 2.8173 ms 59.8%
SingleProcess AUTOTUNE benchmarking takes 2.3086 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x3840)
  triton_mm_11423 0.0385 ms 100.0%
  triton_mm_11428 0.0390 ms 98.8%
  triton_mm_11424 0.0391 ms 98.4%
  triton_mm_11422 0.0401 ms 96.0%
  triton_mm_11437 0.0405 ms 95.0%
  triton_mm_11425 0.0410 ms 94.1%
  triton_mm_11432 0.0410 ms 94.1%
  triton_mm_11435 0.0412 ms 93.6%
  mm 0.0444 ms 86.7%
  triton_mm_11431 0.0446 ms 86.4%
SingleProcess AUTOTUNE benchmarking takes 1.8785 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x32768, 6864x1920, 1920x32768)
  triton_mm_11492 7.0308 ms 100.0%
  bias_addmm 7.1090 ms 98.9%
  addmm 7.8713 ms 89.3%
  triton_mm_11489 8.6481 ms 81.3%
  triton_mm_11493 8.7171 ms 80.7%
  triton_mm_11485 8.8437 ms 79.5%
  triton_mm_11490 9.3529 ms 75.2%
  triton_mm_11487 9.5484 ms 73.6%
  triton_mm_11486 9.6114 ms 73.2%
  triton_mm_11482 11.7356 ms 59.9%
SingleProcess AUTOTUNE benchmarking takes 3.2804 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x768, 6864x2048, 2048x768)
  triton_mm_11549 0.1813 ms 100.0%
  bias_addmm 0.1835 ms 98.8%
  addmm 0.2084 ms 87.0%
  triton_mm_11550 0.2201 ms 82.4%
  triton_mm_11546 0.2218 ms 81.7%
  triton_mm_11542 0.2243 ms 80.9%
  triton_mm_11543 0.2352 ms 77.1%
  triton_mm_11547 0.2383 ms 76.1%
  triton_mm_11544 0.2459 ms 73.8%
  triton_mm_11539 0.3016 ms 60.1%
SingleProcess AUTOTUNE benchmarking takes 2.1316 seconds and 0.0008 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:39<1:54:36, 159.92s/it]AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_8757 1.6527 ms 100.0%
  mm 1.8146 ms 91.1%
  triton_mm_8754 2.0531 ms 80.5%
  triton_mm_8750 2.0801 ms 79.5%
  triton_mm_8758 2.0808 ms 79.4%
  triton_mm_8755 2.2082 ms 74.8%
  triton_mm_8751 2.2550 ms 73.3%
  triton_mm_8752 2.2607 ms 73.1%
  triton_mm_8747 2.7369 ms 60.4%
  triton_mm_8748 2.8593 ms 57.8%
SingleProcess AUTOTUNE benchmarking takes 2.2749 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6411 ms 100.0%
  triton_mm_8850 1.6521 ms 99.3%
  triton_mm_8847 2.0368 ms 80.6%
  triton_mm_8851 2.0729 ms 79.2%
  triton_mm_8843 2.1339 ms 76.9%
  triton_mm_8848 2.2076 ms 74.3%
  triton_mm_8845 2.2536 ms 72.8%
  triton_mm_8844 2.2729 ms 72.2%
  triton_mm_8840 2.7553 ms 59.6%
  triton_mm_8849 2.9021 ms 56.5%
SingleProcess AUTOTUNE benchmarking takes 2.2850 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_9594 1.6531 ms 100.0%
  mm 1.6777 ms 98.5%
  triton_mm_9591 2.0413 ms 81.0%
  triton_mm_9595 2.0763 ms 79.6%
  triton_mm_9587 2.0819 ms 79.4%
  triton_mm_9592 2.2072 ms 74.9%
  triton_mm_9588 2.2544 ms 73.3%
  triton_mm_9589 2.2593 ms 73.2%
  triton_mm_9584 2.7788 ms 59.5%
  triton_mm_9585 2.8509 ms 58.0%
SingleProcess AUTOTUNE benchmarking takes 2.2920 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6426 ms 100.0%
  triton_mm_11268 1.6525 ms 99.4%
  triton_mm_11265 2.0412 ms 80.5%
  triton_mm_11269 2.0693 ms 79.4%
  triton_mm_11261 2.1051 ms 78.0%
  triton_mm_11266 2.2076 ms 74.4%
  triton_mm_11262 2.2553 ms 72.8%
  triton_mm_11263 2.2579 ms 72.7%
  triton_mm_11258 2.7659 ms 59.4%
  triton_mm_11259 2.8576 ms 57.5%
SingleProcess AUTOTUNE benchmarking takes 2.2803 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6431 ms 100.0%
  triton_mm_11361 1.6716 ms 98.3%
  triton_mm_11358 2.0424 ms 80.5%
  triton_mm_11362 2.0758 ms 79.2%
  triton_mm_11354 2.0834 ms 78.9%
  triton_mm_11359 2.2087 ms 74.4%
  triton_mm_11355 2.2555 ms 72.8%
  triton_mm_11356 2.2634 ms 72.6%
  triton_mm_11351 2.7652 ms 59.4%
  triton_mm_11352 2.8513 ms 57.6%
SingleProcess AUTOTUNE benchmarking takes 2.3064 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6360 ms 100.0%
  triton_mm_11454 1.6528 ms 99.0%
  triton_mm_11451 2.0419 ms 80.1%
  triton_mm_11455 2.0750 ms 78.8%
  triton_mm_11447 2.0822 ms 78.6%
  triton_mm_11452 2.2080 ms 74.1%
  triton_mm_11448 2.2546 ms 72.6%
  triton_mm_11449 2.2590 ms 72.4%
  triton_mm_11444 2.7652 ms 59.2%
  triton_mm_11445 2.8600 ms 57.2%
SingleProcess AUTOTUNE benchmarking takes 2.2925 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x2048)
  mm 0.4321 ms 100.0%
  triton_mm_11511 0.4419 ms 97.8%
  triton_mm_11512 0.5513 ms 78.4%
  triton_mm_11508 0.5517 ms 78.3%
  triton_mm_11504 0.5534 ms 78.1%
  triton_mm_11505 0.6015 ms 71.8%
  triton_mm_11509 0.6023 ms 71.7%
  triton_mm_11506 0.6132 ms 70.5%
  triton_mm_11501 0.7469 ms 57.8%
  triton_mm_11502 0.7645 ms 56.5%
SingleProcess AUTOTUNE benchmarking takes 2.0878 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(6864x2048, 2048x2048)
  mm 0.4565 ms 100.0%
  triton_mm_11530 0.4715 ms 96.8%
  triton_mm_11531 0.5826 ms 78.4%
  triton_mm_11527 0.5836 ms 78.2%
  triton_mm_11523 0.5853 ms 78.0%
  triton_mm_11524 0.6249 ms 73.0%
  triton_mm_11528 0.6277 ms 72.7%
  triton_mm_11525 0.6412 ms 71.2%
  triton_mm_11520 0.7907 ms 57.7%
  triton_mm_11521 0.8232 ms 55.5%
SingleProcess AUTOTUNE benchmarking takes 2.1035 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_8663 0.0753 ms 100.0%
  triton_mm_8668 0.0761 ms 99.0%
  triton_mm_8659 0.0765 ms 98.4%
  triton_mm_8662 0.0779 ms 96.7%
  triton_mm_8653 0.0782 ms 96.3%
  triton_mm_8654 0.0783 ms 96.2%
  mm 0.0785 ms 96.0%
  triton_mm_8666 0.0785 ms 95.9%
  triton_mm_8665 0.0792 ms 95.1%
  triton_mm_8661 0.0795 ms 94.7%
SingleProcess AUTOTUNE benchmarking takes 1.9098 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(6840x28, 28x1920)
  triton_mm_8675 0.0391 ms 100.0%
  triton_mm_8680 0.0392 ms 99.9%
  triton_mm_8678 0.0392 ms 99.8%
  triton_mm_8677 0.0393 ms 99.6%
  triton_mm_8676 0.0403 ms 97.1%
  triton_mm_8681 0.0409 ms 95.6%
  triton_mm_8673 0.0424 ms 92.2%
  triton_mm_8684 0.0429 ms 91.3%
  triton_mm_8672 0.0435 ms 89.9%
  triton_mm_8671 0.0439 ms 89.1%
SingleProcess AUTOTUNE benchmarking takes 2.1218 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8702 1.2660 ms 100.0%
  bias_addmm 1.2841 ms 98.6%
  addmm 1.4586 ms 86.8%
  triton_mm_8699 1.5493 ms 81.7%
  triton_mm_8695 1.5711 ms 80.6%
  triton_mm_8703 1.5783 ms 80.2%
  triton_mm_8700 1.6915 ms 74.8%
  triton_mm_8696 1.6995 ms 74.5%
  triton_mm_8697 1.7129 ms 73.9%
  triton_mm_8692 2.0833 ms 60.8%
SingleProcess AUTOTUNE benchmarking takes 2.3626 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8721 0.4339 ms 100.0%
  mm 0.4477 ms 96.9%
  triton_mm_8714 0.5172 ms 83.9%
  triton_mm_8718 0.5180 ms 83.8%
  triton_mm_8722 0.5325 ms 81.5%
  triton_mm_8715 0.5571 ms 77.9%
  triton_mm_8719 0.5585 ms 77.7%
  triton_mm_8716 0.5704 ms 76.1%
  triton_mm_8711 0.6945 ms 62.5%
  triton_mm_8712 0.7198 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.0774 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8776 1.6621 ms 100.0%
  mm 1.6871 ms 98.5%
  triton_mm_8777 1.9743 ms 84.2%
  triton_mm_8773 2.1271 ms 78.1%
  triton_mm_8769 2.1524 ms 77.2%
  triton_mm_8770 2.2052 ms 75.4%
  triton_mm_8774 2.2095 ms 75.2%
  triton_mm_8771 2.2251 ms 74.7%
  triton_mm_8766 2.8045 ms 59.3%
  triton_mm_8767 2.8513 ms 58.3%
SingleProcess AUTOTUNE benchmarking takes 2.3053 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_9014 0.0753 ms 100.0%
  triton_mm_9010 0.0766 ms 98.3%
  triton_mm_9019 0.0766 ms 98.3%
  triton_mm_9005 0.0774 ms 97.2%
  triton_mm_9013 0.0785 ms 95.9%
  triton_mm_9004 0.0787 ms 95.6%
  mm 0.0788 ms 95.6%
  triton_mm_9012 0.0789 ms 95.4%
  triton_mm_9017 0.0790 ms 95.3%
  triton_mm_9016 0.0798 ms 94.3%
SingleProcess AUTOTUNE benchmarking takes 1.8930 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9055 1.6612 ms 100.0%
  mm 1.7152 ms 96.9%
  triton_mm_9056 1.9753 ms 84.1%
  triton_mm_9048 2.1297 ms 78.0%
  triton_mm_9052 2.1413 ms 77.6%
  triton_mm_9049 2.1974 ms 75.6%
  triton_mm_9050 2.2061 ms 75.3%
  triton_mm_9053 2.2190 ms 74.9%
  triton_mm_9045 2.8026 ms 59.3%
  triton_mm_9046 2.8464 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 2.3044 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_9074 1.2658 ms 100.0%
  bias_addmm 1.2844 ms 98.6%
  addmm 1.4577 ms 86.8%
  triton_mm_9071 1.5402 ms 82.2%
  triton_mm_9067 1.5692 ms 80.7%
  triton_mm_9075 1.5796 ms 80.1%
  triton_mm_9072 1.6781 ms 75.4%
  triton_mm_9069 1.7053 ms 74.2%
  triton_mm_9068 1.7155 ms 73.8%
  triton_mm_9064 2.0814 ms 60.8%
SingleProcess AUTOTUNE benchmarking takes 2.3593 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9093 0.4333 ms 100.0%
  mm 0.4467 ms 97.0%
  triton_mm_9086 0.5157 ms 84.0%
  triton_mm_9090 0.5176 ms 83.7%
  triton_mm_9094 0.5323 ms 81.4%
  triton_mm_9087 0.5557 ms 78.0%
  triton_mm_9091 0.5588 ms 77.6%
  triton_mm_9088 0.5716 ms 75.8%
  triton_mm_9083 0.6988 ms 62.0%
  triton_mm_9084 0.7176 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.0843 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_10209 0.4333 ms 100.0%
  mm 0.4456 ms 97.2%
  triton_mm_10202 0.5172 ms 83.8%
  triton_mm_10206 0.5191 ms 83.5%
  triton_mm_10210 0.5325 ms 81.4%
  triton_mm_10203 0.5575 ms 77.7%
  triton_mm_10207 0.5615 ms 77.2%
  triton_mm_10204 0.5723 ms 75.7%
  triton_mm_10199 0.7005 ms 61.9%
  triton_mm_10200 0.7278 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.0881 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_11139 0.4330 ms 100.0%
  mm 0.4451 ms 97.3%
  triton_mm_11132 0.5165 ms 83.8%
  triton_mm_11136 0.5193 ms 83.4%
  triton_mm_11140 0.5338 ms 81.1%
  triton_mm_11133 0.5567 ms 77.8%
  triton_mm_11137 0.5609 ms 77.2%
  triton_mm_11134 0.5709 ms 75.8%
  triton_mm_11129 0.6979 ms 62.0%
  triton_mm_11130 0.7180 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.0764 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_11153 0.0753 ms 100.0%
  triton_mm_11158 0.0760 ms 99.0%
  triton_mm_11149 0.0760 ms 99.0%
  triton_mm_11144 0.0783 ms 96.2%
  triton_mm_11152 0.0783 ms 96.2%
  triton_mm_11156 0.0784 ms 95.9%
  triton_mm_11143 0.0788 ms 95.5%
  triton_mm_11155 0.0792 ms 95.1%
  triton_mm_11151 0.0794 ms 94.8%
  mm 0.0795 ms 94.7%
SingleProcess AUTOTUNE benchmarking takes 1.9019 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_11194 1.6625 ms 100.0%
  mm 1.7177 ms 96.8%
  triton_mm_11195 1.9750 ms 84.2%
  triton_mm_11187 2.1277 ms 78.1%
  triton_mm_11191 2.1439 ms 77.5%
  triton_mm_11188 2.2075 ms 75.3%
  triton_mm_11192 2.2117 ms 75.2%
  triton_mm_11189 2.2262 ms 74.7%
  triton_mm_11184 2.7788 ms 59.8%
  triton_mm_11185 2.8522 ms 58.3%
SingleProcess AUTOTUNE benchmarking takes 2.3067 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_11213 1.2774 ms 100.0%
  bias_addmm 1.2982 ms 98.4%
  addmm 1.4585 ms 87.6%
  triton_mm_11210 1.5238 ms 83.8%
  triton_mm_11206 1.5693 ms 81.4%
  triton_mm_11214 1.5932 ms 80.2%
  triton_mm_11211 1.6769 ms 76.2%
  triton_mm_11208 1.7067 ms 74.8%
  triton_mm_11207 1.7152 ms 74.5%
  triton_mm_11203 2.0821 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.3619 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_11246 0.0755 ms 100.0%
  triton_mm_11242 0.0761 ms 99.2%
  triton_mm_11251 0.0767 ms 98.5%
  triton_mm_11245 0.0777 ms 97.2%
  triton_mm_11236 0.0781 ms 96.6%
  triton_mm_11237 0.0784 ms 96.3%
  triton_mm_11249 0.0785 ms 96.2%
  triton_mm_11248 0.0791 ms 95.4%
  mm 0.0796 ms 94.9%
  triton_mm_11244 0.0797 ms 94.7%
SingleProcess AUTOTUNE benchmarking takes 1.9013 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_11287 1.6970 ms 100.0%
  mm 1.7172 ms 98.8%
  triton_mm_11288 1.9830 ms 85.6%
  triton_mm_11280 2.1022 ms 80.7%
  triton_mm_11284 2.1177 ms 80.1%
  triton_mm_11281 2.1802 ms 77.8%
  triton_mm_11285 2.1908 ms 77.5%
  triton_mm_11282 2.2202 ms 76.4%
  triton_mm_11277 2.7833 ms 61.0%
  triton_mm_11278 2.8510 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.3024 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_11306 1.2657 ms 100.0%
  bias_addmm 1.2849 ms 98.5%
  addmm 1.4590 ms 86.8%
  triton_mm_11303 1.5237 ms 83.1%
  triton_mm_11307 1.5740 ms 80.4%
  triton_mm_11299 1.5877 ms 79.7%
  triton_mm_11304 1.6784 ms 75.4%
  triton_mm_11301 1.7060 ms 74.2%
  triton_mm_11300 1.7328 ms 73.0%
  triton_mm_11296 2.0783 ms 60.9%
SingleProcess AUTOTUNE benchmarking takes 2.3679 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x3840)
  triton_mm_11423 0.0398 ms 100.0%
  triton_mm_11428 0.0399 ms 99.6%
  triton_mm_11424 0.0400 ms 99.4%
  triton_mm_11437 0.0406 ms 98.0%
  triton_mm_11432 0.0408 ms 97.4%
  triton_mm_11422 0.0409 ms 97.3%
  triton_mm_11425 0.0410 ms 97.1%
  triton_mm_11435 0.0412 ms 96.5%
  mm 0.0446 ms 89.2%
  triton_mm_11431 0.0446 ms 89.1%
SingleProcess AUTOTUNE benchmarking takes 1.8746 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x32768, 6864x1920, 1920x32768)
  triton_mm_11492 7.0582 ms 100.0%
  bias_addmm 7.1094 ms 99.3%
  addmm 7.9966 ms 88.3%
  triton_mm_11489 8.6526 ms 81.6%
  triton_mm_11493 8.7574 ms 80.6%
  triton_mm_11485 8.8442 ms 79.8%
  triton_mm_11490 9.4300 ms 74.8%
  triton_mm_11486 9.5228 ms 74.1%
  triton_mm_11487 9.6234 ms 73.3%
  triton_mm_11482 11.7787 ms 59.9%
SingleProcess AUTOTUNE benchmarking takes 3.2985 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(6864x768, 6864x2048, 2048x768)
  triton_mm_11549 0.1815 ms 100.0%
  bias_addmm 0.1840 ms 98.6%
  addmm 0.2087 ms 87.0%
  triton_mm_11546 0.2206 ms 82.3%
  triton_mm_11550 0.2211 ms 82.1%
  triton_mm_11542 0.2214 ms 82.0%
  triton_mm_11543 0.2337 ms 77.7%
  triton_mm_11547 0.2349 ms 77.3%
  triton_mm_11544 0.2458 ms 73.9%
  triton_mm_11539 0.3009 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.1372 seconds and 0.0009 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:39<1:54:37, 159.93s/it]AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_8757 1.6860 ms 100.0%
  mm 1.8113 ms 93.1%
  triton_mm_8754 2.0363 ms 82.8%
  triton_mm_8758 2.0902 ms 80.7%
  triton_mm_8750 2.1617 ms 78.0%
  triton_mm_8755 2.2256 ms 75.8%
  triton_mm_8752 2.2581 ms 74.7%
  triton_mm_8751 2.2888 ms 73.7%
  triton_mm_8747 2.7543 ms 61.2%
  triton_mm_8748 2.8784 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2940 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6444 ms 100.0%
  triton_mm_8850 1.6706 ms 98.4%
  triton_mm_8847 2.0586 ms 79.9%
  triton_mm_8851 2.0867 ms 78.8%
  triton_mm_8843 2.1370 ms 77.0%
  triton_mm_8848 2.2428 ms 73.3%
  triton_mm_8844 2.2536 ms 73.0%
  triton_mm_8845 2.2906 ms 71.8%
  triton_mm_8840 2.7863 ms 59.0%
  triton_mm_8849 2.9045 ms 56.6%
SingleProcess AUTOTUNE benchmarking takes 2.3041 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6596 ms 100.0%
  triton_mm_9036 1.6703 ms 99.4%
  triton_mm_9033 2.0588 ms 80.6%
  triton_mm_9037 2.0843 ms 79.6%
  triton_mm_9029 2.1120 ms 78.6%
  triton_mm_9034 2.2296 ms 74.4%
  triton_mm_9030 2.2529 ms 73.7%
  triton_mm_9031 2.3084 ms 71.9%
  triton_mm_9026 2.7858 ms 59.6%
  triton_mm_9035 2.9035 ms 57.2%
SingleProcess AUTOTUNE benchmarking takes 2.3012 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6573 ms 100.0%
  triton_mm_9129 1.6669 ms 99.4%
  triton_mm_9126 2.0482 ms 80.9%
  triton_mm_9130 2.0781 ms 79.8%
  triton_mm_9122 2.1628 ms 76.6%
  triton_mm_9127 2.2291 ms 74.3%
  triton_mm_9123 2.2527 ms 73.6%
  triton_mm_9124 2.2910 ms 72.3%
  triton_mm_9119 2.7860 ms 59.5%
  triton_mm_9128 2.9028 ms 57.1%
SingleProcess AUTOTUNE benchmarking takes 2.3049 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6587 ms 100.0%
  triton_mm_9315 1.6706 ms 99.3%
  triton_mm_9312 2.0364 ms 81.4%
  triton_mm_9316 2.0927 ms 79.3%
  triton_mm_9308 2.1387 ms 77.6%
  triton_mm_9313 2.2294 ms 74.4%
  triton_mm_9309 2.2806 ms 72.7%
  triton_mm_9310 2.2850 ms 72.6%
  triton_mm_9305 2.7839 ms 59.6%
  triton_mm_9314 2.9037 ms 57.1%
SingleProcess AUTOTUNE benchmarking takes 2.3064 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6601 ms 100.0%
  triton_mm_10896 1.6716 ms 99.3%
  triton_mm_10893 2.0364 ms 81.5%
  triton_mm_10897 2.0816 ms 79.7%
  triton_mm_10889 2.1647 ms 76.7%
  triton_mm_10894 2.2288 ms 74.5%
  triton_mm_10890 2.2526 ms 73.7%
  triton_mm_10891 2.2829 ms 72.7%
  triton_mm_10886 2.7562 ms 60.2%
  triton_mm_10895 2.9053 ms 57.1%
SingleProcess AUTOTUNE benchmarking takes 2.3060 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x2048)
  mm 0.4379 ms 100.0%
  triton_mm_11511 0.4499 ms 97.3%
  triton_mm_11504 0.5552 ms 78.9%
  triton_mm_11512 0.5558 ms 78.8%
  triton_mm_11508 0.5581 ms 78.5%
  triton_mm_11509 0.5997 ms 73.0%
  triton_mm_11505 0.6007 ms 72.9%
  triton_mm_11506 0.6192 ms 70.7%
  triton_mm_11501 0.7552 ms 58.0%
  triton_mm_11502 0.7751 ms 56.5%
SingleProcess AUTOTUNE benchmarking takes 2.0992 seconds and 0.0013 seconds precompiling
AUTOTUNE mm(6864x2048, 2048x2048)
  mm 0.4605 ms 100.0%
  triton_mm_11530 0.4730 ms 97.3%
  triton_mm_11531 0.5840 ms 78.9%
  triton_mm_11527 0.5894 ms 78.1%
  triton_mm_11523 0.5954 ms 77.3%
  triton_mm_11524 0.6310 ms 73.0%
  triton_mm_11528 0.6333 ms 72.7%
  triton_mm_11525 0.6443 ms 71.5%
  triton_mm_11520 0.7941 ms 58.0%
  triton_mm_11521 0.8197 ms 56.2%
SingleProcess AUTOTUNE benchmarking takes 2.1051 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_8663 0.0748 ms 100.0%
  triton_mm_8659 0.0756 ms 98.9%
  triton_mm_8668 0.0757 ms 98.8%
  triton_mm_8662 0.0769 ms 97.2%
  triton_mm_8666 0.0771 ms 97.0%
  triton_mm_8654 0.0774 ms 96.7%
  triton_mm_8661 0.0775 ms 96.5%
  triton_mm_8665 0.0784 ms 95.4%
  mm 0.0785 ms 95.3%
  triton_mm_8656 0.0794 ms 94.3%
SingleProcess AUTOTUNE benchmarking takes 1.9219 seconds and 0.0021 seconds precompiling
AUTOTUNE mm(6840x28, 28x1920)
  triton_mm_8680 0.0390 ms 100.0%
  triton_mm_8676 0.0391 ms 99.6%
  triton_mm_8677 0.0392 ms 99.5%
  triton_mm_8678 0.0400 ms 97.4%
  triton_mm_8675 0.0401 ms 97.2%
  triton_mm_8681 0.0406 ms 95.9%
  triton_mm_8673 0.0416 ms 93.7%
  triton_mm_8672 0.0436 ms 89.4%
  triton_mm_8684 0.0440 ms 88.5%
  triton_mm_8671 0.0449 ms 86.8%
SingleProcess AUTOTUNE benchmarking takes 2.1492 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8702 1.2640 ms 100.0%
  bias_addmm 1.2907 ms 97.9%
  addmm 1.4607 ms 86.5%
  triton_mm_8699 1.5361 ms 82.3%
  triton_mm_8695 1.5727 ms 80.4%
  triton_mm_8703 1.5801 ms 80.0%
  triton_mm_8700 1.6745 ms 75.5%
  triton_mm_8696 1.7150 ms 73.7%
  triton_mm_8697 1.7198 ms 73.5%
  triton_mm_8692 2.0960 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.3593 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8721 0.4387 ms 100.0%
  mm 0.4517 ms 97.1%
  triton_mm_8714 0.5183 ms 84.6%
  triton_mm_8718 0.5228 ms 83.9%
  triton_mm_8722 0.5383 ms 81.5%
  triton_mm_8715 0.5565 ms 78.8%
  triton_mm_8719 0.5625 ms 78.0%
  triton_mm_8716 0.5721 ms 76.7%
  triton_mm_8711 0.7000 ms 62.7%
  triton_mm_8712 0.7251 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0931 seconds and 0.0011 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8776 1.6840 ms 100.0%
  mm 1.7086 ms 98.6%
  triton_mm_8777 1.9774 ms 85.2%
  triton_mm_8773 2.1360 ms 78.8%
  triton_mm_8769 2.1565 ms 78.1%
  triton_mm_8774 2.2151 ms 76.0%
  triton_mm_8771 2.2216 ms 75.8%
  triton_mm_8770 2.2264 ms 75.6%
  triton_mm_8766 2.7945 ms 60.3%
  triton_mm_8767 2.8744 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2916 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(24x1920, 1920x3840)
  triton_mm_11423 0.0383 ms 100.0%
  triton_mm_11424 0.0391 ms 98.0%
  triton_mm_11428 0.0396 ms 96.7%
  triton_mm_11437 0.0399 ms 96.1%
  triton_mm_11432 0.0406 ms 94.3%
  triton_mm_11425 0.0408 ms 94.0%
  triton_mm_11422 0.0409 ms 93.7%
  triton_mm_11435 0.0409 ms 93.7%
  triton_mm_11431 0.0442 ms 86.7%
  mm 0.0447 ms 85.6%
SingleProcess AUTOTUNE benchmarking takes 1.8991 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(6864x32768, 6864x1920, 1920x32768)
  triton_mm_11492 7.0399 ms 100.0%
  bias_addmm 7.0740 ms 99.5%
  addmm 8.0049 ms 87.9%
  triton_mm_11489 8.6217 ms 81.7%
  triton_mm_11493 8.8018 ms 80.0%
  triton_mm_11485 8.9392 ms 78.8%
  triton_mm_11490 9.4151 ms 74.8%
  triton_mm_11487 9.6708 ms 72.8%
  triton_mm_11486 9.6854 ms 72.7%
  triton_mm_11482 11.9796 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 3.2952 seconds and 0.0011 seconds precompiling
AUTOTUNE addmm(6864x768, 6864x2048, 2048x768)
  triton_mm_11549 0.1839 ms 100.0%
  bias_addmm 0.1858 ms 98.9%
  addmm 0.2110 ms 87.1%
  triton_mm_11550 0.2217 ms 82.9%
  triton_mm_11546 0.2247 ms 81.8%
  triton_mm_11542 0.2250 ms 81.7%
  triton_mm_11543 0.2383 ms 77.1%
  triton_mm_11547 0.2389 ms 77.0%
  triton_mm_11544 0.2457 ms 74.8%
  triton_mm_11539 0.3028 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.1514 seconds and 0.0011 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:39<1:54:37, 159.93s/it]AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_8757 1.6709 ms 100.0%
  mm 1.8130 ms 92.2%
  triton_mm_8754 2.0151 ms 82.9%
  triton_mm_8750 2.0646 ms 80.9%
  triton_mm_8758 2.0761 ms 80.5%
  triton_mm_8755 2.2080 ms 75.7%
  triton_mm_8751 2.2292 ms 75.0%
  triton_mm_8752 2.2531 ms 74.2%
  triton_mm_8747 2.7600 ms 60.5%
  triton_mm_8748 2.8555 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2735 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6420 ms 100.0%
  triton_mm_8850 1.6561 ms 99.1%
  triton_mm_8847 2.0144 ms 81.5%
  triton_mm_8851 2.0744 ms 79.2%
  triton_mm_8843 2.0882 ms 78.6%
  triton_mm_8848 2.2073 ms 74.4%
  triton_mm_8845 2.2534 ms 72.9%
  triton_mm_8844 2.2548 ms 72.8%
  triton_mm_8840 2.7593 ms 59.5%
  triton_mm_8841 2.8557 ms 57.5%
SingleProcess AUTOTUNE benchmarking takes 2.2779 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6438 ms 100.0%
  triton_mm_9408 1.6510 ms 99.6%
  triton_mm_9405 2.0340 ms 80.8%
  triton_mm_9409 2.0615 ms 79.7%
  triton_mm_9401 2.0831 ms 78.9%
  triton_mm_9406 2.2025 ms 74.6%
  triton_mm_9402 2.2501 ms 73.1%
  triton_mm_9403 2.2688 ms 72.5%
  triton_mm_9398 2.7529 ms 59.7%
  triton_mm_9399 2.8477 ms 57.7%
SingleProcess AUTOTUNE benchmarking takes 2.2887 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6423 ms 100.0%
  triton_mm_9687 1.6520 ms 99.4%
  triton_mm_9684 2.0325 ms 80.8%
  triton_mm_9688 2.0735 ms 79.2%
  triton_mm_9680 2.1338 ms 77.0%
  triton_mm_9685 2.2015 ms 74.6%
  triton_mm_9682 2.2443 ms 73.2%
  triton_mm_9681 2.2499 ms 73.0%
  triton_mm_9677 2.7519 ms 59.7%
  triton_mm_9678 2.8777 ms 57.1%
SingleProcess AUTOTUNE benchmarking takes 2.2810 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6183 ms 100.0%
  triton_mm_9873 1.6690 ms 97.0%
  triton_mm_9870 2.0387 ms 79.4%
  triton_mm_9866 2.0897 ms 77.4%
  triton_mm_9874 2.0918 ms 77.4%
  triton_mm_9871 2.2080 ms 73.3%
  triton_mm_9867 2.2304 ms 72.6%
  triton_mm_9868 2.2564 ms 71.7%
  triton_mm_9863 2.7614 ms 58.6%
  triton_mm_9864 2.8574 ms 56.6%
SingleProcess AUTOTUNE benchmarking takes 2.2877 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x2048)
  mm 0.4340 ms 100.0%
  triton_mm_11511 0.4474 ms 97.0%
  triton_mm_11512 0.5506 ms 78.8%
  triton_mm_11508 0.5543 ms 78.3%
  triton_mm_11504 0.5572 ms 77.9%
  triton_mm_11509 0.5942 ms 73.0%
  triton_mm_11505 0.5949 ms 72.9%
  triton_mm_11506 0.6093 ms 71.2%
  triton_mm_11501 0.7437 ms 58.3%
  triton_mm_11502 0.7662 ms 56.6%
SingleProcess AUTOTUNE benchmarking takes 2.0803 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x2048, 2048x2048)
  mm 0.4596 ms 100.0%
  triton_mm_11530 0.4726 ms 97.3%
  triton_mm_11531 0.5816 ms 79.0%
  triton_mm_11527 0.5847 ms 78.6%
  triton_mm_11523 0.5929 ms 77.5%
  triton_mm_11524 0.6268 ms 73.3%
  triton_mm_11528 0.6282 ms 73.2%
  triton_mm_11525 0.6385 ms 72.0%
  triton_mm_11520 0.7875 ms 58.4%
  triton_mm_11521 0.8211 ms 56.0%
SingleProcess AUTOTUNE benchmarking takes 2.0982 seconds and 0.0012 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_8663 0.0752 ms 100.0%
  triton_mm_8668 0.0771 ms 97.5%
  triton_mm_8659 0.0772 ms 97.3%
  triton_mm_8662 0.0782 ms 96.1%
  triton_mm_8654 0.0786 ms 95.6%
  mm 0.0787 ms 95.5%
  triton_mm_8653 0.0793 ms 94.8%
  triton_mm_8661 0.0794 ms 94.7%
  triton_mm_8666 0.0796 ms 94.5%
  triton_mm_8665 0.0800 ms 93.9%
SingleProcess AUTOTUNE benchmarking takes 1.9125 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(6840x28, 28x1920)
  triton_mm_8675 0.0392 ms 100.0%
  triton_mm_8678 0.0406 ms 96.7%
  triton_mm_8680 0.0406 ms 96.6%
  triton_mm_8676 0.0407 ms 96.5%
  triton_mm_8677 0.0407 ms 96.4%
  triton_mm_8681 0.0411 ms 95.5%
  triton_mm_8673 0.0429 ms 91.5%
  triton_mm_8672 0.0437 ms 89.8%
  triton_mm_8684 0.0444 ms 88.5%
  triton_mm_8671 0.0454 ms 86.3%
SingleProcess AUTOTUNE benchmarking takes 2.1201 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8702 1.2557 ms 100.0%
  bias_addmm 1.2853 ms 97.7%
  addmm 1.4612 ms 85.9%
  triton_mm_8699 1.5203 ms 82.6%
  triton_mm_8695 1.5740 ms 79.8%
  triton_mm_8703 1.5768 ms 79.6%
  triton_mm_8700 1.6770 ms 74.9%
  triton_mm_8696 1.6982 ms 73.9%
  triton_mm_8697 1.7160 ms 73.2%
  triton_mm_8692 2.0973 ms 59.9%
SingleProcess AUTOTUNE benchmarking takes 2.3612 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8721 0.4387 ms 100.0%
  mm 0.4512 ms 97.2%
  triton_mm_8714 0.5171 ms 84.8%
  triton_mm_8718 0.5195 ms 84.4%
  triton_mm_8722 0.5360 ms 81.8%
  triton_mm_8715 0.5559 ms 78.9%
  triton_mm_8719 0.5581 ms 78.6%
  triton_mm_8716 0.5717 ms 76.7%
  triton_mm_8711 0.6961 ms 63.0%
  triton_mm_8712 0.7226 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0775 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8776 1.6866 ms 100.0%
  mm 1.6952 ms 99.5%
  triton_mm_8777 1.9809 ms 85.1%
  triton_mm_8773 2.1172 ms 79.7%
  triton_mm_8769 2.1235 ms 79.4%
  triton_mm_8774 2.1714 ms 77.7%
  triton_mm_8770 2.1787 ms 77.4%
  triton_mm_8771 2.1899 ms 77.0%
  triton_mm_8766 2.7941 ms 60.4%
  triton_mm_8767 2.8721 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.3008 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_9479 0.0756 ms 100.0%
  triton_mm_9484 0.0771 ms 98.1%
  triton_mm_9475 0.0773 ms 97.9%
  triton_mm_9478 0.0776 ms 97.5%
  triton_mm_9470 0.0779 ms 97.1%
  triton_mm_9477 0.0783 ms 96.6%
  triton_mm_9469 0.0785 ms 96.4%
  triton_mm_9482 0.0794 ms 95.3%
  mm 0.0795 ms 95.1%
  triton_mm_9472 0.0801 ms 94.4%
SingleProcess AUTOTUNE benchmarking takes 1.9051 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9520 1.6860 ms 100.0%
  mm 1.6939 ms 99.5%
  triton_mm_9521 1.9809 ms 85.1%
  triton_mm_9513 2.1034 ms 80.2%
  triton_mm_9517 2.1198 ms 79.5%
  triton_mm_9518 2.1779 ms 77.4%
  triton_mm_9514 2.1842 ms 77.2%
  triton_mm_9515 2.1901 ms 77.0%
  triton_mm_9510 2.7977 ms 60.3%
  triton_mm_9511 2.8456 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2901 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_9539 1.2571 ms 100.0%
  bias_addmm 1.2952 ms 97.1%
  addmm 1.4562 ms 86.3%
  triton_mm_9536 1.5198 ms 82.7%
  triton_mm_9532 1.5573 ms 80.7%
  triton_mm_9540 1.5764 ms 79.7%
  triton_mm_9537 1.6784 ms 74.9%
  triton_mm_9533 1.6993 ms 74.0%
  triton_mm_9534 1.7154 ms 73.3%
  triton_mm_9529 2.0988 ms 59.9%
SingleProcess AUTOTUNE benchmarking takes 2.3672 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9558 0.4368 ms 100.0%
  mm 0.4527 ms 96.5%
  triton_mm_9551 0.5178 ms 84.4%
  triton_mm_9555 0.5195 ms 84.1%
  triton_mm_9559 0.5358 ms 81.5%
  triton_mm_9552 0.5562 ms 78.5%
  triton_mm_9556 0.5596 ms 78.1%
  triton_mm_9553 0.5718 ms 76.4%
  triton_mm_9548 0.6974 ms 62.6%
  triton_mm_9549 0.7226 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.0795 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_9632 1.2665 ms 100.0%
  bias_addmm 1.2960 ms 97.7%
  addmm 1.4627 ms 86.6%
  triton_mm_9629 1.5201 ms 83.3%
  triton_mm_9625 1.5740 ms 80.5%
  triton_mm_9633 1.5753 ms 80.4%
  triton_mm_9630 1.6778 ms 75.5%
  triton_mm_9626 1.6993 ms 74.5%
  triton_mm_9627 1.7157 ms 73.8%
  triton_mm_9622 2.0975 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.3614 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9651 0.4387 ms 100.0%
  mm 0.4524 ms 97.0%
  triton_mm_9644 0.5181 ms 84.7%
  triton_mm_9648 0.5200 ms 84.4%
  triton_mm_9652 0.5367 ms 81.7%
  triton_mm_9645 0.5566 ms 78.8%
  triton_mm_9649 0.5587 ms 78.5%
  triton_mm_9646 0.5734 ms 76.5%
  triton_mm_9641 0.6982 ms 62.8%
  triton_mm_9642 0.7262 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.0804 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x3840)
  triton_mm_11428 0.0386 ms 100.0%
  triton_mm_11424 0.0394 ms 98.0%
  triton_mm_11423 0.0396 ms 97.6%
  triton_mm_11432 0.0396 ms 97.5%
  triton_mm_11437 0.0401 ms 96.3%
  triton_mm_11422 0.0412 ms 93.8%
  triton_mm_11425 0.0413 ms 93.5%
  triton_mm_11435 0.0420 ms 92.0%
  triton_mm_11431 0.0432 ms 89.4%
  mm 0.0444 ms 86.8%
SingleProcess AUTOTUNE benchmarking takes 1.8743 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(6864x32768, 6864x1920, 1920x32768)
  triton_mm_11492 7.0564 ms 100.0%
  bias_addmm 7.0829 ms 99.6%
  addmm 7.9885 ms 88.3%
  triton_mm_11489 8.6455 ms 81.6%
  triton_mm_11493 8.7931 ms 80.2%
  triton_mm_11485 8.8535 ms 79.7%
  triton_mm_11490 9.4422 ms 74.7%
  triton_mm_11487 9.5704 ms 73.7%
  triton_mm_11486 9.6412 ms 73.2%
  triton_mm_11482 11.7398 ms 60.1%
SingleProcess AUTOTUNE benchmarking takes 3.2773 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x768, 6864x2048, 2048x768)
  triton_mm_11549 0.1832 ms 100.0%
  bias_addmm 0.1853 ms 98.8%
  addmm 0.2104 ms 87.1%
  triton_mm_11550 0.2200 ms 83.2%
  triton_mm_11542 0.2218 ms 82.6%
  triton_mm_11546 0.2251 ms 81.4%
  triton_mm_11543 0.2357 ms 77.7%
  triton_mm_11547 0.2395 ms 76.5%
  triton_mm_11544 0.2414 ms 75.9%
  triton_mm_11539 0.3023 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.1304 seconds and 0.0009 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:39<1:54:37, 159.94s/it]AUTOTUNE mm(6864x1920, 1920x7680)
  triton_mm_8757 1.6512 ms 100.0%
  mm 1.8182 ms 90.8%
  triton_mm_8754 2.0341 ms 81.2%
  triton_mm_8750 2.0830 ms 79.3%
  triton_mm_8758 2.0881 ms 79.1%
  triton_mm_8755 2.2021 ms 75.0%
  triton_mm_8752 2.2502 ms 73.4%
  triton_mm_8751 2.2515 ms 73.3%
  triton_mm_8747 2.7524 ms 60.0%
  triton_mm_8748 2.8472 ms 58.0%
SingleProcess AUTOTUNE benchmarking takes 2.2805 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6443 ms 100.0%
  triton_mm_8850 1.6792 ms 97.9%
  triton_mm_8847 2.0359 ms 80.8%
  triton_mm_8843 2.0820 ms 79.0%
  triton_mm_8851 2.0937 ms 78.5%
  triton_mm_8848 2.2027 ms 74.7%
  triton_mm_8844 2.2512 ms 73.0%
  triton_mm_8845 2.2718 ms 72.4%
  triton_mm_8840 2.7508 ms 59.8%
  triton_mm_8841 2.8798 ms 57.1%
SingleProcess AUTOTUNE benchmarking takes 2.2801 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x7680)
  mm 1.6422 ms 100.0%
  triton_mm_10896 1.6528 ms 99.4%
  triton_mm_10893 2.0426 ms 80.4%
  triton_mm_10897 2.0891 ms 78.6%
  triton_mm_10889 2.0894 ms 78.6%
  triton_mm_10894 2.2300 ms 73.6%
  triton_mm_10890 2.2579 ms 72.7%
  triton_mm_10891 2.2746 ms 72.2%
  triton_mm_10886 2.7595 ms 59.5%
  triton_mm_10887 2.8888 ms 56.8%
SingleProcess AUTOTUNE benchmarking takes 2.2882 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x2048)
  mm 0.4360 ms 100.0%
  triton_mm_11511 0.4472 ms 97.5%
  triton_mm_11512 0.5544 ms 78.6%
  triton_mm_11508 0.5573 ms 78.2%
  triton_mm_11504 0.5588 ms 78.0%
  triton_mm_11505 0.5965 ms 73.1%
  triton_mm_11509 0.5969 ms 73.0%
  triton_mm_11506 0.6154 ms 70.8%
  triton_mm_11501 0.7503 ms 58.1%
  triton_mm_11502 0.7740 ms 56.3%
SingleProcess AUTOTUNE benchmarking takes 2.0905 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(6864x2048, 2048x2048)
  mm 0.4605 ms 100.0%
  triton_mm_11530 0.4738 ms 97.2%
  triton_mm_11531 0.5865 ms 78.5%
  triton_mm_11527 0.5896 ms 78.1%
  triton_mm_11523 0.5951 ms 77.4%
  triton_mm_11524 0.6294 ms 73.2%
  triton_mm_11528 0.6295 ms 73.2%
  triton_mm_11525 0.6427 ms 71.6%
  triton_mm_11520 0.7923 ms 58.1%
  triton_mm_11521 0.8201 ms 56.1%
SingleProcess AUTOTUNE benchmarking takes 2.0878 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(24x1920, 1920x11520)
  triton_mm_8663 0.0743 ms 100.0%
  triton_mm_8668 0.0762 ms 97.6%
  triton_mm_8659 0.0764 ms 97.2%
  triton_mm_8662 0.0770 ms 96.5%
  triton_mm_8661 0.0771 ms 96.4%
  triton_mm_8654 0.0773 ms 96.1%
  triton_mm_8666 0.0780 ms 95.3%
  mm 0.0780 ms 95.3%
  triton_mm_8665 0.0785 ms 94.6%
  triton_mm_8653 0.0790 ms 94.0%
SingleProcess AUTOTUNE benchmarking takes 1.9073 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6840x28, 28x1920)
  triton_mm_8678 0.0393 ms 100.0%
  triton_mm_8680 0.0393 ms 100.0%
  triton_mm_8677 0.0394 ms 99.8%
  triton_mm_8675 0.0395 ms 99.6%
  triton_mm_8676 0.0407 ms 96.7%
  triton_mm_8681 0.0412 ms 95.5%
  triton_mm_8673 0.0428 ms 91.7%
  triton_mm_8684 0.0433 ms 90.8%
  triton_mm_8672 0.0450 ms 87.4%
  triton_mm_8671 0.0453 ms 86.8%
SingleProcess AUTOTUNE benchmarking takes 2.1298 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8702 1.2629 ms 100.0%
  bias_addmm 1.2851 ms 98.3%
  addmm 1.4599 ms 86.5%
  triton_mm_8699 1.5343 ms 82.3%
  triton_mm_8695 1.5705 ms 80.4%
  triton_mm_8703 1.5732 ms 80.3%
  triton_mm_8700 1.6703 ms 75.6%
  triton_mm_8696 1.6967 ms 74.4%
  triton_mm_8697 1.7183 ms 73.5%
  triton_mm_8692 2.0905 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.3571 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8721 0.4378 ms 100.0%
  mm 0.4518 ms 96.9%
  triton_mm_8714 0.5187 ms 84.4%
  triton_mm_8718 0.5199 ms 84.2%
  triton_mm_8722 0.5357 ms 81.7%
  triton_mm_8715 0.5576 ms 78.5%
  triton_mm_8719 0.5612 ms 78.0%
  triton_mm_8716 0.5718 ms 76.6%
  triton_mm_8711 0.6982 ms 62.7%
  triton_mm_8712 0.7240 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0731 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8776 1.6861 ms 100.0%
  mm 1.6990 ms 99.2%
  triton_mm_8777 1.9808 ms 85.1%
  triton_mm_8769 2.1366 ms 78.9%
  triton_mm_8773 2.1650 ms 77.9%
  triton_mm_8774 2.1820 ms 77.3%
  triton_mm_8770 2.1853 ms 77.2%
  triton_mm_8771 2.2033 ms 76.5%
  triton_mm_8766 2.7973 ms 60.3%
  triton_mm_8767 2.8412 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2799 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_8869 1.6862 ms 100.0%
  mm 1.7042 ms 98.9%
  triton_mm_8870 1.9945 ms 84.5%
  triton_mm_8866 2.1441 ms 78.6%
  triton_mm_8862 2.1619 ms 78.0%
  triton_mm_8867 2.1814 ms 77.3%
  triton_mm_8863 2.2067 ms 76.4%
  triton_mm_8864 2.2232 ms 75.8%
  triton_mm_8859 2.7949 ms 60.3%
  triton_mm_8860 2.8406 ms 59.4%
SingleProcess AUTOTUNE benchmarking takes 2.2881 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_8888 1.2661 ms 100.0%
  bias_addmm 1.2971 ms 97.6%
  addmm 1.4648 ms 86.4%
  triton_mm_8885 1.5363 ms 82.4%
  triton_mm_8881 1.5714 ms 80.6%
  triton_mm_8889 1.5936 ms 79.5%
  triton_mm_8886 1.6708 ms 75.8%
  triton_mm_8882 1.7160 ms 73.8%
  triton_mm_8883 1.7165 ms 73.8%
  triton_mm_8878 2.0907 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.3448 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_8907 0.4386 ms 100.0%
  mm 0.4526 ms 96.9%
  triton_mm_8900 0.5190 ms 84.5%
  triton_mm_8904 0.5204 ms 84.3%
  triton_mm_8908 0.5358 ms 81.8%
  triton_mm_8901 0.5592 ms 78.4%
  triton_mm_8905 0.5610 ms 78.2%
  triton_mm_8902 0.5733 ms 76.5%
  triton_mm_8897 0.6984 ms 62.8%
  triton_mm_8898 0.7244 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.0730 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9148 1.6875 ms 100.0%
  mm 1.6990 ms 99.3%
  triton_mm_9149 1.9950 ms 84.6%
  triton_mm_9145 2.1454 ms 78.7%
  triton_mm_9141 2.1607 ms 78.1%
  triton_mm_9146 2.1812 ms 77.4%
  triton_mm_9142 2.2083 ms 76.4%
  triton_mm_9143 2.2206 ms 76.0%
  triton_mm_9138 2.7977 ms 60.3%
  triton_mm_9139 2.8408 ms 59.4%
SingleProcess AUTOTUNE benchmarking takes 2.2986 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_9167 1.2641 ms 100.0%
  bias_addmm 1.2980 ms 97.4%
  addmm 1.4660 ms 86.2%
  triton_mm_9164 1.5358 ms 82.3%
  triton_mm_9160 1.5706 ms 80.5%
  triton_mm_9168 1.5898 ms 79.5%
  triton_mm_9165 1.6709 ms 75.7%
  triton_mm_9161 1.7161 ms 73.7%
  triton_mm_9162 1.7186 ms 73.6%
  triton_mm_9157 2.0909 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.3531 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9186 0.4387 ms 100.0%
  mm 0.4528 ms 96.9%
  triton_mm_9179 0.5205 ms 84.3%
  triton_mm_9183 0.5234 ms 83.8%
  triton_mm_9187 0.5389 ms 81.4%
  triton_mm_9180 0.5621 ms 78.0%
  triton_mm_9184 0.5646 ms 77.7%
  triton_mm_9181 0.5771 ms 76.0%
  triton_mm_9176 0.7042 ms 62.3%
  triton_mm_9177 0.7241 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.0775 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9520 1.6881 ms 100.0%
  mm 1.7056 ms 99.0%
  triton_mm_9521 1.9980 ms 84.5%
  triton_mm_9513 2.1414 ms 78.8%
  triton_mm_9517 2.1462 ms 78.7%
  triton_mm_9518 2.1829 ms 77.3%
  triton_mm_9514 2.1917 ms 77.0%
  triton_mm_9515 2.2324 ms 75.6%
  triton_mm_9510 2.8030 ms 60.2%
  triton_mm_9511 2.8493 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2818 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_9539 1.2652 ms 100.0%
  bias_addmm 1.3052 ms 96.9%
  addmm 1.4674 ms 86.2%
  triton_mm_9536 1.5391 ms 82.2%
  triton_mm_9532 1.5744 ms 80.4%
  triton_mm_9540 1.5892 ms 79.6%
  triton_mm_9537 1.6972 ms 74.6%
  triton_mm_9533 1.7210 ms 73.5%
  triton_mm_9534 1.7229 ms 73.4%
  triton_mm_9529 2.0991 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.3545 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9558 0.4396 ms 100.0%
  mm 0.4548 ms 96.7%
  triton_mm_9551 0.5199 ms 84.6%
  triton_mm_9555 0.5242 ms 83.9%
  triton_mm_9559 0.5399 ms 81.4%
  triton_mm_9552 0.5612 ms 78.3%
  triton_mm_9556 0.5654 ms 77.8%
  triton_mm_9553 0.5770 ms 76.2%
  triton_mm_9548 0.6983 ms 63.0%
  triton_mm_9549 0.7254 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.0857 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x7680, 7680x1920)
  triton_mm_9706 1.6874 ms 100.0%
  mm 1.7014 ms 99.2%
  triton_mm_9707 1.9956 ms 84.6%
  triton_mm_9703 2.1474 ms 78.6%
  triton_mm_9699 2.1628 ms 78.0%
  triton_mm_9704 2.1821 ms 77.3%
  triton_mm_9700 2.2087 ms 76.4%
  triton_mm_9701 2.2264 ms 75.8%
  triton_mm_9696 2.8241 ms 59.7%
  triton_mm_9697 2.8754 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2908 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(6864x5760, 6864x1920, 1920x5760)
  triton_mm_9725 1.2651 ms 100.0%
  bias_addmm 1.3049 ms 97.0%
  addmm 1.4673 ms 86.2%
  triton_mm_9722 1.5347 ms 82.4%
  triton_mm_9718 1.5706 ms 80.5%
  triton_mm_9726 1.5890 ms 79.6%
  triton_mm_9723 1.6712 ms 75.7%
  triton_mm_9719 1.7169 ms 73.7%
  triton_mm_9720 1.7202 ms 73.5%
  triton_mm_9715 2.1195 ms 59.7%
SingleProcess AUTOTUNE benchmarking takes 2.3446 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(6864x1920, 1920x1920)
  triton_mm_9744 0.4396 ms 100.0%
  mm 0.4538 ms 96.9%
  triton_mm_9737 0.5220 ms 84.2%
  triton_mm_9741 0.5232 ms 84.0%
  triton_mm_9745 0.5387 ms 81.6%
  triton_mm_9738 0.5630 ms 78.1%
  triton_mm_9742 0.5652 ms 77.8%
  triton_mm_9739 0.5773 ms 76.2%
  triton_mm_9734 0.7049 ms 62.4%
  triton_mm_9735 0.7243 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0818 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(24x1920, 1920x3840)
  triton_mm_11423 0.0386 ms 100.0%
  triton_mm_11428 0.0390 ms 98.9%
  triton_mm_11424 0.0393 ms 98.0%
  triton_mm_11432 0.0398 ms 96.9%
  triton_mm_11437 0.0407 ms 94.7%
  triton_mm_11422 0.0412 ms 93.6%
  triton_mm_11425 0.0418 ms 92.3%
  triton_mm_11435 0.0421 ms 91.5%
  triton_mm_11431 0.0433 ms 89.1%
  mm 0.0448 ms 86.1%
SingleProcess AUTOTUNE benchmarking takes 1.8785 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(6864x32768, 6864x1920, 1920x32768)
  triton_mm_11492 7.0292 ms 100.0%
  bias_addmm 7.0632 ms 99.5%
  addmm 7.9259 ms 88.7%
  triton_mm_11489 8.7343 ms 80.5%
  triton_mm_11493 8.7946 ms 79.9%
  triton_mm_11485 8.8414 ms 79.5%
  triton_mm_11490 9.5590 ms 73.5%
  triton_mm_11486 9.6110 ms 73.1%
  triton_mm_11487 9.6293 ms 73.0%
  triton_mm_11482 11.9951 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 3.2822 seconds and 0.0009 seconds precompiling
AUTOTUNE addmm(6864x768, 6864x2048, 2048x768)
  triton_mm_11549 0.1848 ms 100.0%
  bias_addmm 0.1860 ms 99.3%
  addmm 0.2125 ms 87.0%
  triton_mm_11550 0.2227 ms 83.0%
  triton_mm_11546 0.2256 ms 81.9%
  triton_mm_11542 0.2261 ms 81.7%
  triton_mm_11543 0.2377 ms 77.7%
  triton_mm_11547 0.2404 ms 76.9%
  triton_mm_11544 0.2452 ms 75.4%
  triton_mm_11539 0.3038 ms 60.8%
SingleProcess AUTOTUNE benchmarking takes 2.1414 seconds and 0.0009 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:40<1:54:41, 160.05s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:43<47:45, 68.23s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:43<47:45, 68.23s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:43<47:45, 68.24s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:43<47:45, 68.23s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:43<47:45, 68.24s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:44<47:47, 68.28s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:45<25:46, 37.72s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:45<25:46, 37.72s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:45<25:46, 37.72s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:45<25:46, 37.72s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:45<25:46, 37.72s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:45<25:47, 37.75s/it]
Validation Progress:   9%|9         | 4/44 [02:46<15:33, 23.35s/it]
Validation Progress:   9%|9         | 4/44 [02:46<15:33, 23.35s/it]
Validation Progress:   9%|9         | 4/44 [02:46<15:34, 23.35s/it]
Validation Progress:   9%|9         | 4/44 [02:46<15:34, 23.35s/it]
Validation Progress:   9%|9         | 4/44 [02:46<15:34, 23.35s/it]
Validation Progress:   9%|9         | 4/44 [02:46<15:34, 23.37s/it]
Validation Progress:  11%|#1        | 5/44 [02:48<10:04, 15.49s/it]
Validation Progress:  11%|#1        | 5/44 [02:48<10:04, 15.49s/it]
Validation Progress:  11%|#1        | 5/44 [02:48<10:04, 15.49s/it]
Validation Progress:  11%|#1        | 5/44 [02:48<10:04, 15.49s/it]
Validation Progress:  11%|#1        | 5/44 [02:48<10:04, 15.49s/it]
Validation Progress:  11%|#1        | 5/44 [02:48<10:04, 15.50s/it]
Validation Progress:  14%|#3        | 6/44 [02:49<06:46, 10.69s/it]
Validation Progress:  14%|#3        | 6/44 [02:49<06:46, 10.69s/it]
Validation Progress:  14%|#3        | 6/44 [02:49<06:46, 10.69s/it]
Validation Progress:  14%|#3        | 6/44 [02:49<06:46, 10.70s/it]
Validation Progress:  14%|#3        | 6/44 [02:49<06:46, 10.70s/it]
Validation Progress:  14%|#3        | 6/44 [02:49<06:46, 10.70s/it]
Validation Progress:  16%|#5        | 7/44 [02:51<04:45,  7.72s/it]
Validation Progress:  16%|#5        | 7/44 [02:51<04:45,  7.72s/it]
Validation Progress:  16%|#5        | 7/44 [02:51<04:45,  7.72s/it]
Validation Progress:  16%|#5        | 7/44 [02:51<04:45,  7.72s/it]
Validation Progress:  16%|#5        | 7/44 [02:51<04:45,  7.72s/it]
Validation Progress:  16%|#5        | 7/44 [02:51<04:45,  7.72s/it]
Validation Progress:  18%|#8        | 8/44 [02:52<03:26,  5.75s/it]
Validation Progress:  18%|#8        | 8/44 [02:52<03:26,  5.75s/it]
Validation Progress:  18%|#8        | 8/44 [02:52<03:26,  5.75s/it]
Validation Progress:  18%|#8        | 8/44 [02:52<03:26,  5.75s/it]
Validation Progress:  18%|#8        | 8/44 [02:52<03:26,  5.75s/it]
Validation Progress:  18%|#8        | 8/44 [02:52<03:26,  5.75s/it]
Validation Progress:  20%|##        | 9/44 [02:54<02:32,  4.36s/it]
Validation Progress:  20%|##        | 9/44 [02:54<02:32,  4.36s/it]
Validation Progress:  20%|##        | 9/44 [02:54<02:32,  4.36s/it]
Validation Progress:  20%|##        | 9/44 [02:54<02:32,  4.36s/it]
Validation Progress:  20%|##        | 9/44 [02:54<02:32,  4.36s/it]
Validation Progress:  20%|##        | 9/44 [02:54<02:32,  4.36s/it]
Validation Progress:  23%|##2       | 10/44 [02:55<01:56,  3.44s/it]
Validation Progress:  23%|##2       | 10/44 [02:55<01:56,  3.44s/it]
Validation Progress:  23%|##2       | 10/44 [02:55<01:56,  3.44s/it]
Validation Progress:  23%|##2       | 10/44 [02:55<01:56,  3.44s/it]
Validation Progress:  23%|##2       | 10/44 [02:55<01:56,  3.44s/it]
Validation Progress:  23%|##2       | 10/44 [02:55<01:56,  3.44s/it]
Validation Progress:  25%|##5       | 11/44 [02:56<01:33,  2.82s/it]
Validation Progress:  25%|##5       | 11/44 [02:57<01:34,  2.86s/it]
Validation Progress:  25%|##5       | 11/44 [02:57<01:34,  2.86s/it]
Validation Progress:  25%|##5       | 11/44 [02:57<01:34,  2.86s/it]
Validation Progress:  25%|##5       | 11/44 [02:57<01:34,  2.86s/it]
Validation Progress:  25%|##5       | 11/44 [02:57<01:34,  2.86s/it]
Validation Progress:  27%|##7       | 12/44 [02:58<01:18,  2.44s/it]
Validation Progress:  27%|##7       | 12/44 [02:58<01:18,  2.46s/it]
Validation Progress:  27%|##7       | 12/44 [02:58<01:18,  2.46s/it]
Validation Progress:  27%|##7       | 12/44 [02:58<01:18,  2.46s/it]
Validation Progress:  27%|##7       | 12/44 [02:58<01:18,  2.46s/it]
Validation Progress:  27%|##7       | 12/44 [02:58<01:18,  2.47s/it]
Validation Progress:  30%|##9       | 13/44 [03:00<01:08,  2.20s/it]
Validation Progress:  30%|##9       | 13/44 [03:00<01:08,  2.22s/it]
Validation Progress:  30%|##9       | 13/44 [03:00<01:08,  2.22s/it]
Validation Progress:  30%|##9       | 13/44 [03:00<01:08,  2.22s/it]
Validation Progress:  30%|##9       | 13/44 [03:00<01:08,  2.22s/it]
Validation Progress:  30%|##9       | 13/44 [03:00<01:08,  2.22s/it]
Validation Progress:  32%|###1      | 14/44 [03:01<00:59,  1.99s/it]
Validation Progress:  32%|###1      | 14/44 [03:01<00:59,  1.99s/it]
Validation Progress:  32%|###1      | 14/44 [03:01<00:59,  1.99s/it]
Validation Progress:  32%|###1      | 14/44 [03:01<00:59,  1.99s/it]
Validation Progress:  32%|###1      | 14/44 [03:01<00:59,  1.98s/it]
Validation Progress:  32%|###1      | 14/44 [03:01<00:59,  1.99s/it]
Validation Progress:  34%|###4      | 15/44 [03:03<00:53,  1.84s/it]
Validation Progress:  34%|###4      | 15/44 [03:03<00:53,  1.84s/it]
Validation Progress:  34%|###4      | 15/44 [03:03<00:53,  1.84s/it]
Validation Progress:  34%|###4      | 15/44 [03:03<00:53,  1.83s/it]
Validation Progress:  34%|###4      | 15/44 [03:03<00:53,  1.84s/it]
Validation Progress:  34%|###4      | 15/44 [03:03<00:53,  1.84s/it]
Validation Progress:  36%|###6      | 16/44 [03:04<00:48,  1.74s/it]
Validation Progress:  36%|###6      | 16/44 [03:04<00:49,  1.75s/it]
Validation Progress:  36%|###6      | 16/44 [03:04<00:49,  1.75s/it]
Validation Progress:  36%|###6      | 16/44 [03:04<00:49,  1.75s/it]
Validation Progress:  36%|###6      | 16/44 [03:04<00:49,  1.75s/it]
Validation Progress:  36%|###6      | 16/44 [03:04<00:49,  1.75s/it]
Validation Progress:  39%|###8      | 17/44 [03:06<00:45,  1.67s/it]
Validation Progress:  39%|###8      | 17/44 [03:06<00:45,  1.68s/it]
Validation Progress:  39%|###8      | 17/44 [03:06<00:45,  1.68s/it]
Validation Progress:  39%|###8      | 17/44 [03:06<00:45,  1.68s/it]
Validation Progress:  39%|###8      | 17/44 [03:06<00:45,  1.68s/it]
Validation Progress:  39%|###8      | 17/44 [03:06<00:45,  1.68s/it]
Validation Progress:  41%|####      | 18/44 [03:07<00:41,  1.61s/it]
Validation Progress:  41%|####      | 18/44 [03:07<00:41,  1.61s/it]
Validation Progress:  41%|####      | 18/44 [03:07<00:41,  1.61s/it]
Validation Progress:  41%|####      | 18/44 [03:07<00:41,  1.61s/it]
Validation Progress:  41%|####      | 18/44 [03:07<00:41,  1.61s/it]
Validation Progress:  41%|####      | 18/44 [03:07<00:41,  1.61s/it]
Validation Progress:  43%|####3     | 19/44 [03:09<00:40,  1.61s/it]
Validation Progress:  43%|####3     | 19/44 [03:09<00:40,  1.61s/it]
Validation Progress:  43%|####3     | 19/44 [03:09<00:40,  1.61s/it]
Validation Progress:  43%|####3     | 19/44 [03:09<00:40,  1.61s/it]
Validation Progress:  43%|####3     | 19/44 [03:09<00:40,  1.61s/it]
Validation Progress:  43%|####3     | 19/44 [03:09<00:40,  1.61s/it]
Validation Progress:  45%|####5     | 20/44 [03:10<00:37,  1.58s/it]
Validation Progress:  45%|####5     | 20/44 [03:10<00:37,  1.58s/it]
Validation Progress:  45%|####5     | 20/44 [03:10<00:37,  1.58s/it]
Validation Progress:  45%|####5     | 20/44 [03:10<00:37,  1.58s/it]
Validation Progress:  45%|####5     | 20/44 [03:10<00:37,  1.58s/it]
Validation Progress:  45%|####5     | 20/44 [03:10<00:37,  1.58s/it]
Validation Progress:  48%|####7     | 21/44 [03:12<00:35,  1.55s/it]
Validation Progress:  48%|####7     | 21/44 [03:12<00:35,  1.55s/it]
Validation Progress:  48%|####7     | 21/44 [03:12<00:35,  1.55s/it]
Validation Progress:  48%|####7     | 21/44 [03:12<00:35,  1.55s/it]
Validation Progress:  48%|####7     | 21/44 [03:12<00:35,  1.55s/it]
Validation Progress:  48%|####7     | 21/44 [03:12<00:35,  1.55s/it]
Validation Progress:  50%|#####     | 22/44 [03:13<00:34,  1.57s/it]
Validation Progress:  50%|#####     | 22/44 [03:13<00:34,  1.57s/it]
Validation Progress:  50%|#####     | 22/44 [03:13<00:34,  1.57s/it]
Validation Progress:  50%|#####     | 22/44 [03:13<00:34,  1.57s/it]
Validation Progress:  50%|#####     | 22/44 [03:13<00:34,  1.57s/it]
Validation Progress:  50%|#####     | 22/44 [03:13<00:34,  1.57s/it]
Validation Progress:  52%|#####2    | 23/44 [03:15<00:32,  1.55s/it]
Validation Progress:  52%|#####2    | 23/44 [03:15<00:32,  1.55s/it]
Validation Progress:  52%|#####2    | 23/44 [03:15<00:32,  1.55s/it]
Validation Progress:  52%|#####2    | 23/44 [03:15<00:32,  1.55s/it]
Validation Progress:  52%|#####2    | 23/44 [03:15<00:32,  1.55s/it]
Validation Progress:  52%|#####2    | 23/44 [03:15<00:32,  1.55s/it]
Validation Progress:  55%|#####4    | 24/44 [03:16<00:31,  1.55s/it]
Validation Progress:  55%|#####4    | 24/44 [03:16<00:31,  1.55s/it]
Validation Progress:  55%|#####4    | 24/44 [03:16<00:31,  1.55s/it]
Validation Progress:  55%|#####4    | 24/44 [03:16<00:31,  1.55s/it]
Validation Progress:  55%|#####4    | 24/44 [03:16<00:31,  1.55s/it]
Validation Progress:  55%|#####4    | 24/44 [03:16<00:31,  1.55s/it]
Validation Progress:  57%|#####6    | 25/44 [03:18<00:28,  1.52s/it]
Validation Progress:  57%|#####6    | 25/44 [03:18<00:28,  1.52s/it]
Validation Progress:  57%|#####6    | 25/44 [03:18<00:28,  1.52s/it]
Validation Progress:  57%|#####6    | 25/44 [03:18<00:28,  1.52s/it]
Validation Progress:  57%|#####6    | 25/44 [03:18<00:28,  1.52s/it]
Validation Progress:  57%|#####6    | 25/44 [03:18<00:28,  1.52s/it]
Validation Progress:  59%|#####9    | 26/44 [03:19<00:26,  1.50s/it]
Validation Progress:  59%|#####9    | 26/44 [03:19<00:26,  1.50s/it]
Validation Progress:  59%|#####9    | 26/44 [03:19<00:26,  1.50s/it]
Validation Progress:  59%|#####9    | 26/44 [03:19<00:26,  1.50s/it]
Validation Progress:  59%|#####9    | 26/44 [03:19<00:26,  1.50s/it]
Validation Progress:  59%|#####9    | 26/44 [03:19<00:26,  1.50s/it]
Validation Progress:  61%|######1   | 27/44 [03:21<00:25,  1.49s/it]
Validation Progress:  61%|######1   | 27/44 [03:21<00:25,  1.49s/it]
Validation Progress:  61%|######1   | 27/44 [03:21<00:25,  1.49s/it]
Validation Progress:  61%|######1   | 27/44 [03:21<00:25,  1.49s/it]
Validation Progress:  61%|######1   | 27/44 [03:21<00:25,  1.49s/it]
Validation Progress:  61%|######1   | 27/44 [03:21<00:25,  1.49s/it]
Validation Progress:  64%|######3   | 28/44 [03:22<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [03:22<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [03:22<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [03:22<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [03:22<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [03:22<00:23,  1.48s/it]
Validation Progress:  66%|######5   | 29/44 [03:24<00:22,  1.48s/it]
Validation Progress:  66%|######5   | 29/44 [03:24<00:22,  1.48s/it]
Validation Progress:  66%|######5   | 29/44 [03:24<00:22,  1.48s/it]
Validation Progress:  66%|######5   | 29/44 [03:24<00:22,  1.48s/it]
Validation Progress:  66%|######5   | 29/44 [03:24<00:22,  1.48s/it]
Validation Progress:  66%|######5   | 29/44 [03:24<00:22,  1.48s/it]
Validation Progress:  68%|######8   | 30/44 [03:25<00:20,  1.48s/it]
Validation Progress:  68%|######8   | 30/44 [03:25<00:20,  1.48s/it]
Validation Progress:  68%|######8   | 30/44 [03:25<00:20,  1.48s/it]
Validation Progress:  68%|######8   | 30/44 [03:25<00:20,  1.48s/it]
Validation Progress:  68%|######8   | 30/44 [03:25<00:20,  1.48s/it]
Validation Progress:  68%|######8   | 30/44 [03:25<00:20,  1.48s/it]
Validation Progress:  70%|#######   | 31/44 [03:27<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [03:27<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [03:27<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [03:27<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [03:27<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [03:27<00:19,  1.50s/it]
Validation Progress:  73%|#######2  | 32/44 [03:28<00:18,  1.50s/it]
Validation Progress:  73%|#######2  | 32/44 [03:28<00:18,  1.50s/it]
Validation Progress:  73%|#######2  | 32/44 [03:28<00:18,  1.50s/it]
Validation Progress:  73%|#######2  | 32/44 [03:28<00:18,  1.50s/it]
Validation Progress:  73%|#######2  | 32/44 [03:28<00:18,  1.51s/it]
Validation Progress:  73%|#######2  | 32/44 [03:28<00:18,  1.50s/it]
Validation Progress:  75%|#######5  | 33/44 [03:30<00:17,  1.56s/it]
Validation Progress:  75%|#######5  | 33/44 [03:30<00:17,  1.56s/it]
Validation Progress:  75%|#######5  | 33/44 [03:30<00:17,  1.56s/it]
Validation Progress:  75%|#######5  | 33/44 [03:30<00:17,  1.56s/it]
Validation Progress:  75%|#######5  | 33/44 [03:30<00:17,  1.56s/it]
Validation Progress:  75%|#######5  | 33/44 [03:30<00:17,  1.56s/it]
Validation Progress:  77%|#######7  | 34/44 [03:32<00:15,  1.56s/it]
Validation Progress:  77%|#######7  | 34/44 [03:32<00:15,  1.56s/it]
Validation Progress:  77%|#######7  | 34/44 [03:32<00:15,  1.56s/it]
Validation Progress:  77%|#######7  | 34/44 [03:32<00:15,  1.56s/it]
Validation Progress:  77%|#######7  | 34/44 [03:32<00:15,  1.56s/it]
Validation Progress:  77%|#######7  | 34/44 [03:32<00:15,  1.56s/it]
Validation Progress:  80%|#######9  | 35/44 [03:33<00:14,  1.60s/it]
Validation Progress:  80%|#######9  | 35/44 [03:33<00:14,  1.60s/it]
Validation Progress:  80%|#######9  | 35/44 [03:33<00:14,  1.60s/it]
Validation Progress:  80%|#######9  | 35/44 [03:33<00:14,  1.60s/it]
Validation Progress:  80%|#######9  | 35/44 [03:33<00:14,  1.60s/it]
Validation Progress:  80%|#######9  | 35/44 [03:33<00:14,  1.60s/it]
Validation Progress:  82%|########1 | 36/44 [03:35<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [03:35<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [03:35<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [03:35<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [03:35<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [03:35<00:12,  1.56s/it]
Validation Progress:  84%|########4 | 37/44 [03:36<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [03:36<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [03:36<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [03:36<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [03:36<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [03:36<00:10,  1.53s/it]
Validation Progress:  86%|########6 | 38/44 [03:38<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [03:38<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [03:38<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [03:38<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [03:38<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [03:38<00:09,  1.50s/it]
Validation Progress:  89%|########8 | 39/44 [03:39<00:07,  1.56s/it]
Validation Progress:  89%|########8 | 39/44 [03:39<00:07,  1.56s/it]
Validation Progress:  89%|########8 | 39/44 [03:39<00:07,  1.56s/it]
Validation Progress:  89%|########8 | 39/44 [03:39<00:07,  1.56s/it]
Validation Progress:  89%|########8 | 39/44 [03:39<00:07,  1.56s/it]
Validation Progress:  89%|########8 | 39/44 [03:39<00:07,  1.56s/it]
Validation Progress:  91%|######### | 40/44 [03:41<00:06,  1.58s/it]
Validation Progress:  91%|######### | 40/44 [03:41<00:06,  1.58s/it]
Validation Progress:  91%|######### | 40/44 [03:41<00:06,  1.58s/it]
Validation Progress:  91%|######### | 40/44 [03:41<00:06,  1.58s/it]
Validation Progress:  91%|######### | 40/44 [03:41<00:06,  1.58s/it]
Validation Progress:  91%|######### | 40/44 [03:41<00:06,  1.58s/it]
Validation Progress:  93%|#########3| 41/44 [03:42<00:04,  1.53s/it]
Validation Progress:  93%|#########3| 41/44 [03:42<00:04,  1.53s/it]
Validation Progress:  93%|#########3| 41/44 [03:42<00:04,  1.53s/it]
Validation Progress:  93%|#########3| 41/44 [03:42<00:04,  1.53s/it]
Validation Progress:  93%|#########3| 41/44 [03:42<00:04,  1.53s/it]
Validation Progress:  93%|#########3| 41/44 [03:42<00:04,  1.54s/it]
Validation Progress:  95%|#########5| 42/44 [03:44<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [03:44<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [03:44<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [03:44<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [03:44<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [03:44<00:03,  1.52s/it]
Validation Progress:  98%|#########7| 43/44 [03:45<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [03:45<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [03:45<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [03:45<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [03:45<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [03:45<00:01,  1.51s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  1.35s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  5.15s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  5.15s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  1.36s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  5.15s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  5.15s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  5.15s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  1.37s/it]
Validation Progress: 100%|##########| 44/44 [03:46<00:00,  5.16s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.15s/it]
100%|##########| 1/1 [00:03<00:00,  3.15s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|          | 1/157 [00:01<03:05,  1.19s/it]
  2%|1         | 3/157 [00:01<00:54,  2.84it/s]
  3%|3         | 5/157 [00:01<00:30,  4.98it/s]
  4%|4         | 7/157 [00:01<00:21,  7.11it/s]
  6%|5         | 9/157 [00:01<00:16,  9.11it/s]
  7%|7         | 11/157 [00:01<00:13, 10.89it/s]
  8%|8         | 13/157 [00:01<00:11, 12.09it/s]
 10%|9         | 15/157 [00:02<00:10, 13.27it/s]
 11%|#         | 17/157 [00:02<00:09, 14.17it/s]
 12%|#2        | 19/157 [00:02<00:09, 14.77it/s]
 13%|#3        | 21/157 [00:02<00:09, 14.85it/s]
 15%|#4        | 23/157 [00:02<00:08, 15.45it/s]
 16%|#5        | 25/157 [00:02<00:08, 15.81it/s]
 17%|#7        | 27/157 [00:02<00:08, 16.08it/s]
 18%|#8        | 29/157 [00:02<00:07, 16.18it/s]
 20%|#9        | 31/157 [00:02<00:07, 16.43it/s]
 21%|##1       | 33/157 [00:03<00:07, 16.49it/s]
 22%|##2       | 35/157 [00:03<00:07, 16.65it/s]
 24%|##3       | 37/157 [00:03<00:07, 15.94it/s]
 25%|##4       | 39/157 [00:03<00:07, 16.21it/s]
 26%|##6       | 41/157 [00:03<00:07, 16.34it/s]
 27%|##7       | 43/157 [00:03<00:06, 16.53it/s]
 29%|##8       | 45/157 [00:03<00:06, 16.67it/s]
 30%|##9       | 47/157 [00:03<00:06, 16.64it/s]
 31%|###1      | 49/157 [00:04<00:06, 16.61it/s]
 32%|###2      | 51/157 [00:04<00:06, 16.63it/s]
 34%|###3      | 53/157 [00:04<00:06, 16.67it/s]
 35%|###5      | 55/157 [00:04<00:06, 16.66it/s]
 36%|###6      | 57/157 [00:04<00:05, 16.78it/s]
 38%|###7      | 59/157 [00:04<00:05, 16.88it/s]
 39%|###8      | 61/157 [00:04<00:05, 16.94it/s]
 40%|####      | 63/157 [00:04<00:05, 16.99it/s]
 41%|####1     | 65/157 [00:05<00:05, 17.04it/s]
 43%|####2     | 67/157 [00:05<00:05, 17.04it/s]
 44%|####3     | 69/157 [00:05<00:05, 15.54it/s]
 45%|####5     | 71/157 [00:05<00:05, 15.85it/s]
 46%|####6     | 73/157 [00:05<00:05, 16.08it/s]
 48%|####7     | 75/157 [00:05<00:05, 16.30it/s]
 49%|####9     | 77/157 [00:05<00:04, 16.53it/s]
 50%|#####     | 79/157 [00:05<00:04, 16.72it/s]
 52%|#####1    | 81/157 [00:06<00:04, 16.84it/s]
 53%|#####2    | 83/157 [00:06<00:04, 16.92it/s]
 54%|#####4    | 85/157 [00:06<00:04, 16.98it/s]
 55%|#####5    | 87/157 [00:06<00:04, 16.99it/s]
 57%|#####6    | 89/157 [00:06<00:03, 17.03it/s]
 58%|#####7    | 91/157 [00:06<00:03, 16.95it/s]
 59%|#####9    | 93/157 [00:06<00:03, 16.88it/s]
 61%|######    | 95/157 [00:06<00:03, 16.83it/s]
 62%|######1   | 97/157 [00:06<00:03, 16.80it/s]
 63%|######3   | 99/157 [00:07<00:03, 16.78it/s]
 64%|######4   | 101/157 [00:07<00:03, 16.80it/s]
 66%|######5   | 103/157 [00:07<00:03, 16.78it/s]
 67%|######6   | 105/157 [00:07<00:03, 16.76it/s]
 68%|######8   | 107/157 [00:07<00:02, 16.80it/s]
 69%|######9   | 109/157 [00:07<00:02, 16.88it/s]
 71%|#######   | 111/157 [00:07<00:02, 16.92it/s]
 72%|#######1  | 113/157 [00:07<00:02, 16.97it/s]
 73%|#######3  | 115/157 [00:08<00:02, 16.99it/s]
 75%|#######4  | 117/157 [00:08<00:02, 16.98it/s]
 76%|#######5  | 119/157 [00:08<00:02, 16.88it/s]
 77%|#######7  | 121/157 [00:08<00:02, 16.93it/s]
 78%|#######8  | 123/157 [00:08<00:02, 16.96it/s]
 80%|#######9  | 125/157 [00:08<00:01, 16.99it/s]
 81%|########  | 127/157 [00:08<00:01, 17.03it/s]
 82%|########2 | 129/157 [00:08<00:01, 17.08it/s]
 83%|########3 | 131/157 [00:08<00:01, 17.09it/s]
 85%|########4 | 133/157 [00:09<00:01, 14.60it/s]
 86%|########5 | 135/157 [00:09<00:01, 15.25it/s]
 87%|########7 | 137/157 [00:09<00:01, 15.76it/s]
 89%|########8 | 139/157 [00:09<00:01, 16.13it/s]
 90%|########9 | 141/157 [00:09<00:00, 16.41it/s]
 91%|#########1| 143/157 [00:09<00:00, 16.61it/s]
 92%|#########2| 145/157 [00:09<00:00, 16.78it/s]
 94%|#########3| 147/157 [00:09<00:00, 16.87it/s]
 95%|#########4| 149/157 [00:10<00:00, 16.95it/s]
 96%|#########6| 151/157 [00:10<00:00, 17.03it/s]
 97%|#########7| 153/157 [00:10<00:00, 17.12it/s]
 99%|#########8| 155/157 [00:10<00:00, 17.19it/s]
100%|##########| 157/157 [00:11<00:00,  6.35it/s]
100%|##########| 157/157 [00:11<00:00, 14.01it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:40, 19.13it/s]
  1%|          | 4/781 [00:00<00:40, 19.20it/s]
  1%|          | 6/781 [00:00<00:40, 19.24it/s]
  1%|1         | 8/781 [00:00<00:40, 19.27it/s]
  1%|1         | 10/781 [00:00<00:39, 19.33it/s]
  2%|1         | 12/781 [00:00<00:39, 19.31it/s]
  2%|1         | 14/781 [00:00<00:39, 19.31it/s]
  2%|2         | 16/781 [00:00<00:39, 19.13it/s]
  2%|2         | 18/781 [00:00<00:39, 19.17it/s]
  3%|2         | 20/781 [00:01<00:40, 18.91it/s]
  3%|2         | 22/781 [00:01<00:39, 18.98it/s]
  3%|3         | 24/781 [00:01<00:39, 19.03it/s]
  3%|3         | 26/781 [00:01<00:40, 18.87it/s]
  4%|3         | 28/781 [00:01<00:39, 18.92it/s]
  4%|3         | 30/781 [00:01<00:39, 19.04it/s]
  4%|4         | 32/781 [00:01<00:39, 19.12it/s]
  4%|4         | 34/781 [00:01<00:38, 19.23it/s]
  5%|4         | 36/781 [00:01<00:39, 18.73it/s]
  5%|4         | 38/781 [00:01<00:39, 18.74it/s]
  5%|5         | 40/781 [00:02<00:39, 18.79it/s]
  5%|5         | 42/781 [00:02<00:39, 18.81it/s]
  6%|5         | 44/781 [00:02<00:38, 18.93it/s]
  6%|5         | 46/781 [00:02<00:39, 18.79it/s]
  6%|6         | 48/781 [00:02<00:38, 18.86it/s]
  6%|6         | 50/781 [00:02<00:38, 18.82it/s]
  7%|6         | 52/781 [00:02<00:38, 18.88it/s]
  7%|6         | 54/781 [00:02<00:38, 19.06it/s]
  7%|7         | 56/781 [00:02<00:38, 18.88it/s]
  7%|7         | 58/781 [00:03<00:38, 18.84it/s]
  8%|7         | 60/781 [00:03<00:38, 18.90it/s]
  8%|7         | 62/781 [00:03<00:38, 18.66it/s]
  8%|8         | 64/781 [00:03<00:38, 18.66it/s]
  8%|8         | 66/781 [00:03<00:38, 18.66it/s]
  9%|8         | 68/781 [00:03<00:38, 18.64it/s]
  9%|8         | 70/781 [00:03<00:37, 18.76it/s]
  9%|9         | 72/781 [00:03<00:37, 18.89it/s]
  9%|9         | 74/781 [00:03<00:37, 18.84it/s]
 10%|9         | 76/781 [00:04<00:37, 18.91it/s]
 10%|9         | 78/781 [00:04<00:37, 18.83it/s]
 10%|#         | 80/781 [00:04<00:37, 18.76it/s]
 10%|#         | 82/781 [00:04<00:37, 18.56it/s]
 11%|#         | 84/781 [00:04<00:37, 18.71it/s]
 11%|#1        | 86/781 [00:04<00:37, 18.67it/s]
 11%|#1        | 88/781 [00:04<00:36, 18.86it/s]
 12%|#1        | 90/781 [00:04<00:36, 18.93it/s]
 12%|#1        | 92/781 [00:04<00:36, 18.80it/s]
 12%|#2        | 94/781 [00:04<00:36, 18.77it/s]
 12%|#2        | 96/781 [00:05<00:36, 18.91it/s]
 13%|#2        | 98/781 [00:05<00:36, 18.66it/s]
 13%|#2        | 100/781 [00:05<00:36, 18.78it/s]
 13%|#3        | 102/781 [00:05<00:36, 18.61it/s]
 13%|#3        | 104/781 [00:05<00:36, 18.61it/s]
 14%|#3        | 106/781 [00:05<00:36, 18.58it/s]
 14%|#3        | 108/781 [00:05<00:35, 18.73it/s]
 14%|#4        | 110/781 [00:05<00:35, 18.69it/s]
 14%|#4        | 112/781 [00:05<00:35, 18.73it/s]
 15%|#4        | 114/781 [00:06<00:35, 18.70it/s]
 15%|#4        | 116/781 [00:06<00:35, 18.82it/s]
 15%|#5        | 118/781 [00:06<00:35, 18.90it/s]
 15%|#5        | 120/781 [00:06<00:34, 18.95it/s]
 16%|#5        | 122/781 [00:06<00:34, 19.13it/s]
 16%|#5        | 124/781 [00:06<00:34, 19.10it/s]
 16%|#6        | 126/781 [00:06<00:34, 19.07it/s]
 16%|#6        | 128/781 [00:06<00:34, 19.06it/s]
 17%|#6        | 130/781 [00:06<00:34, 19.05it/s]
 17%|#6        | 132/781 [00:06<00:34, 18.88it/s]
 17%|#7        | 134/781 [00:07<00:34, 18.91it/s]
 17%|#7        | 136/781 [00:07<00:33, 19.11it/s]
 18%|#7        | 138/781 [00:07<00:33, 18.96it/s]
 18%|#7        | 140/781 [00:07<00:33, 19.01it/s]
 18%|#8        | 142/781 [00:07<00:33, 19.03it/s]
 18%|#8        | 144/781 [00:07<00:33, 19.18it/s]
 19%|#8        | 146/781 [00:07<00:33, 18.99it/s]
 19%|#8        | 148/781 [00:07<00:33, 19.03it/s]
 19%|#9        | 150/781 [00:07<00:32, 19.22it/s]
 19%|#9        | 152/781 [00:08<00:33, 19.01it/s]
 20%|#9        | 154/781 [00:08<00:32, 19.05it/s]
 20%|#9        | 156/781 [00:08<00:32, 19.23it/s]
 20%|##        | 158/781 [00:08<00:32, 19.05it/s]
 20%|##        | 160/781 [00:08<00:32, 19.07it/s]
 21%|##        | 162/781 [00:08<00:32, 19.25it/s]
 21%|##        | 164/781 [00:08<00:32, 19.06it/s]
 21%|##1       | 166/781 [00:08<00:31, 19.22it/s]
 22%|##1       | 168/781 [00:08<00:31, 19.18it/s]
 22%|##1       | 170/781 [00:08<00:32, 18.97it/s]
 22%|##2       | 172/781 [00:09<00:31, 19.08it/s]
 22%|##2       | 174/781 [00:09<00:31, 19.09it/s]
 23%|##2       | 176/781 [00:09<00:31, 19.08it/s]
 23%|##2       | 178/781 [00:09<00:31, 19.10it/s]
 23%|##3       | 180/781 [00:09<00:31, 18.96it/s]
 23%|##3       | 182/781 [00:09<00:31, 19.00it/s]
 24%|##3       | 184/781 [00:09<00:31, 19.19it/s]
 24%|##3       | 186/781 [00:09<00:31, 19.10it/s]
 24%|##4       | 188/781 [00:09<00:31, 18.78it/s]
 24%|##4       | 190/781 [00:10<00:31, 18.89it/s]
 25%|##4       | 192/781 [00:10<00:31, 18.80it/s]
 25%|##4       | 194/781 [00:10<00:31, 18.72it/s]
 25%|##5       | 196/781 [00:10<00:30, 18.98it/s]
 25%|##5       | 198/781 [00:10<00:30, 18.98it/s]
 26%|##5       | 200/781 [00:10<00:30, 18.87it/s]
 26%|##5       | 202/781 [00:10<00:30, 18.79it/s]
 26%|##6       | 204/781 [00:10<00:31, 18.61it/s]
 26%|##6       | 206/781 [00:10<00:30, 18.60it/s]
 27%|##6       | 208/781 [00:10<00:30, 18.49it/s]
 27%|##6       | 210/781 [00:11<00:30, 18.61it/s]
 27%|##7       | 212/781 [00:11<00:30, 18.72it/s]
 27%|##7       | 214/781 [00:11<00:30, 18.80it/s]
 28%|##7       | 216/781 [00:11<00:30, 18.63it/s]
 28%|##7       | 218/781 [00:11<00:29, 18.88it/s]
 28%|##8       | 220/781 [00:11<00:29, 18.80it/s]
 28%|##8       | 222/781 [00:11<00:29, 18.71it/s]
 29%|##8       | 224/781 [00:11<00:29, 18.72it/s]
 29%|##8       | 226/781 [00:11<00:29, 18.83it/s]
 29%|##9       | 228/781 [00:12<00:29, 18.95it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.84it/s]
 30%|##9       | 232/781 [00:12<00:29, 18.76it/s]
 30%|##9       | 234/781 [00:12<00:28, 19.00it/s]
 30%|###       | 236/781 [00:12<00:28, 19.02it/s]
 30%|###       | 238/781 [00:12<00:28, 18.96it/s]
 31%|###       | 240/781 [00:12<00:28, 19.15it/s]
 31%|###       | 242/781 [00:12<00:28, 19.07it/s]
 31%|###1      | 244/781 [00:12<00:28, 19.12it/s]
 31%|###1      | 246/781 [00:13<00:27, 19.26it/s]
 32%|###1      | 248/781 [00:13<00:27, 19.09it/s]
 32%|###2      | 250/781 [00:13<00:27, 19.26it/s]
 32%|###2      | 252/781 [00:13<00:27, 19.08it/s]
 33%|###2      | 254/781 [00:13<00:27, 19.12it/s]
 33%|###2      | 256/781 [00:13<00:27, 19.15it/s]
 33%|###3      | 258/781 [00:13<00:27, 19.30it/s]
 33%|###3      | 260/781 [00:13<00:27, 19.30it/s]
 34%|###3      | 262/781 [00:14<00:48, 10.67it/s]
 34%|###3      | 264/781 [00:14<00:41, 12.35it/s]
 34%|###4      | 266/781 [00:14<00:37, 13.74it/s]
 34%|###4      | 268/781 [00:14<00:34, 15.03it/s]
 35%|###4      | 270/781 [00:14<00:31, 16.16it/s]
 35%|###4      | 272/781 [00:14<00:30, 16.87it/s]
 35%|###5      | 274/781 [00:14<00:28, 17.50it/s]
 35%|###5      | 276/781 [00:14<00:27, 18.07it/s]
 36%|###5      | 278/781 [00:14<00:27, 18.24it/s]
 36%|###5      | 280/781 [00:15<00:27, 18.52it/s]
 36%|###6      | 282/781 [00:15<00:26, 18.85it/s]
 36%|###6      | 284/781 [00:15<00:26, 18.82it/s]
 37%|###6      | 286/781 [00:15<00:26, 18.92it/s]
 37%|###6      | 288/781 [00:15<00:25, 19.02it/s]
 37%|###7      | 290/781 [00:15<00:25, 19.22it/s]
 37%|###7      | 292/781 [00:15<00:25, 19.03it/s]
 38%|###7      | 294/781 [00:15<00:25, 18.76it/s]
 38%|###7      | 296/781 [00:15<00:25, 19.05it/s]
 38%|###8      | 298/781 [00:15<00:25, 19.10it/s]
 38%|###8      | 300/781 [00:16<00:25, 19.17it/s]
 39%|###8      | 302/781 [00:16<00:25, 19.06it/s]
 39%|###8      | 304/781 [00:16<00:24, 19.28it/s]
 39%|###9      | 306/781 [00:16<00:24, 19.15it/s]
 39%|###9      | 308/781 [00:16<00:24, 19.25it/s]
 40%|###9      | 310/781 [00:16<00:24, 19.29it/s]
 40%|###9      | 312/781 [00:16<00:24, 19.31it/s]
 40%|####      | 314/781 [00:16<00:24, 19.30it/s]
 40%|####      | 316/781 [00:16<00:24, 19.18it/s]
 41%|####      | 318/781 [00:17<00:24, 19.11it/s]
 41%|####      | 320/781 [00:17<00:24, 19.19it/s]
 41%|####1     | 322/781 [00:17<00:24, 18.90it/s]
 41%|####1     | 324/781 [00:17<00:24, 18.87it/s]
 42%|####1     | 326/781 [00:17<00:24, 18.80it/s]
 42%|####1     | 328/781 [00:17<00:24, 18.79it/s]
 42%|####2     | 330/781 [00:17<00:23, 19.05it/s]
 43%|####2     | 332/781 [00:17<00:23, 18.94it/s]
 43%|####2     | 334/781 [00:17<00:23, 19.05it/s]
 43%|####3     | 336/781 [00:17<00:23, 19.24it/s]
 43%|####3     | 338/781 [00:18<00:23, 19.24it/s]
 44%|####3     | 340/781 [00:18<00:23, 19.07it/s]
 44%|####3     | 342/781 [00:18<00:22, 19.26it/s]
 44%|####4     | 344/781 [00:18<00:22, 19.23it/s]
 44%|####4     | 346/781 [00:18<00:22, 18.94it/s]
 45%|####4     | 348/781 [00:18<00:22, 18.86it/s]
 45%|####4     | 350/781 [00:18<00:22, 18.80it/s]
 45%|####5     | 352/781 [00:18<00:22, 18.77it/s]
 45%|####5     | 354/781 [00:18<00:22, 18.89it/s]
 46%|####5     | 356/781 [00:19<00:22, 19.02it/s]
 46%|####5     | 358/781 [00:19<00:22, 19.12it/s]
 46%|####6     | 360/781 [00:19<00:21, 19.15it/s]
 46%|####6     | 362/781 [00:19<00:22, 19.04it/s]
 47%|####6     | 364/781 [00:19<00:21, 19.10it/s]
 47%|####6     | 366/781 [00:19<00:21, 19.10it/s]
 47%|####7     | 368/781 [00:19<00:21, 19.09it/s]
 47%|####7     | 370/781 [00:19<00:21, 19.12it/s]
 48%|####7     | 372/781 [00:19<00:21, 19.17it/s]
 48%|####7     | 374/781 [00:19<00:21, 19.20it/s]
 48%|####8     | 376/781 [00:20<00:21, 19.24it/s]
 48%|####8     | 378/781 [00:20<00:21, 19.07it/s]
 49%|####8     | 380/781 [00:20<00:21, 18.97it/s]
 49%|####8     | 382/781 [00:20<00:20, 19.01it/s]
 49%|####9     | 384/781 [00:20<00:20, 19.13it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.21it/s]
 50%|####9     | 388/781 [00:20<00:20, 19.21it/s]
 50%|####9     | 390/781 [00:20<00:20, 19.19it/s]
 50%|#####     | 392/781 [00:20<00:20, 19.06it/s]
 50%|#####     | 394/781 [00:21<00:20, 19.00it/s]
 51%|#####     | 396/781 [00:21<00:20, 18.77it/s]
 51%|#####     | 398/781 [00:21<00:20, 18.87it/s]
 51%|#####1    | 400/781 [00:21<00:20, 18.56it/s]
 51%|#####1    | 402/781 [00:21<00:20, 18.66it/s]
 52%|#####1    | 404/781 [00:21<00:20, 18.73it/s]
 52%|#####1    | 406/781 [00:21<00:19, 18.87it/s]
 52%|#####2    | 408/781 [00:21<00:19, 18.92it/s]
 52%|#####2    | 410/781 [00:21<00:19, 18.86it/s]
 53%|#####2    | 412/781 [00:21<00:19, 18.95it/s]
 53%|#####3    | 414/781 [00:22<00:19, 19.06it/s]
 53%|#####3    | 416/781 [00:22<00:19, 19.13it/s]
 54%|#####3    | 418/781 [00:22<00:18, 19.20it/s]
 54%|#####3    | 420/781 [00:22<00:18, 19.07it/s]
 54%|#####4    | 422/781 [00:22<00:18, 19.12it/s]
 54%|#####4    | 424/781 [00:22<00:18, 19.30it/s]
 55%|#####4    | 426/781 [00:22<00:18, 19.30it/s]
 55%|#####4    | 428/781 [00:22<00:18, 18.83it/s]
 55%|#####5    | 430/781 [00:22<00:18, 19.06it/s]
 55%|#####5    | 432/781 [00:23<00:18, 19.10it/s]
 56%|#####5    | 434/781 [00:23<00:18, 19.11it/s]
 56%|#####5    | 436/781 [00:23<00:18, 19.12it/s]
 56%|#####6    | 438/781 [00:23<00:17, 19.13it/s]
 56%|#####6    | 440/781 [00:23<00:17, 19.15it/s]
 57%|#####6    | 442/781 [00:23<00:17, 19.17it/s]
 57%|#####6    | 444/781 [00:23<00:17, 19.16it/s]
 57%|#####7    | 446/781 [00:23<00:17, 19.09it/s]
 57%|#####7    | 448/781 [00:23<00:17, 18.98it/s]
 58%|#####7    | 450/781 [00:23<00:17, 18.92it/s]
 58%|#####7    | 452/781 [00:24<00:17, 18.83it/s]
 58%|#####8    | 454/781 [00:24<00:17, 18.79it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.77it/s]
 59%|#####8    | 458/781 [00:24<00:17, 18.73it/s]
 59%|#####8    | 460/781 [00:24<00:17, 18.71it/s]
 59%|#####9    | 462/781 [00:24<00:17, 18.69it/s]
 59%|#####9    | 464/781 [00:24<00:16, 18.68it/s]
 60%|#####9    | 466/781 [00:24<00:16, 18.67it/s]
 60%|#####9    | 468/781 [00:24<00:16, 18.67it/s]
 60%|######    | 470/781 [00:25<00:16, 18.68it/s]
 60%|######    | 472/781 [00:25<00:16, 18.67it/s]
 61%|######    | 474/781 [00:25<00:16, 18.79it/s]
 61%|######    | 476/781 [00:25<00:16, 18.88it/s]
 61%|######1   | 478/781 [00:25<00:15, 18.97it/s]
 61%|######1   | 480/781 [00:25<00:15, 19.03it/s]
 62%|######1   | 482/781 [00:25<00:15, 19.07it/s]
 62%|######1   | 484/781 [00:25<00:15, 19.09it/s]
 62%|######2   | 486/781 [00:25<00:15, 19.09it/s]
 62%|######2   | 488/781 [00:25<00:15, 19.14it/s]
 63%|######2   | 490/781 [00:26<00:15, 19.16it/s]
 63%|######2   | 492/781 [00:26<00:15, 19.17it/s]
 63%|######3   | 494/781 [00:26<00:14, 19.16it/s]
 64%|######3   | 496/781 [00:26<00:14, 19.15it/s]
 64%|######3   | 498/781 [00:26<00:14, 19.17it/s]
 64%|######4   | 500/781 [00:26<00:14, 19.19it/s]
 64%|######4   | 502/781 [00:26<00:14, 19.17it/s]
 65%|######4   | 504/781 [00:26<00:14, 19.16it/s]
 65%|######4   | 506/781 [00:26<00:14, 19.14it/s]
 65%|######5   | 508/781 [00:27<00:14, 19.12it/s]
 65%|######5   | 510/781 [00:27<00:14, 19.07it/s]
 66%|######5   | 512/781 [00:27<00:14, 19.11it/s]
 66%|######5   | 514/781 [00:27<00:13, 19.14it/s]
 66%|######6   | 516/781 [00:27<00:13, 19.15it/s]
 66%|######6   | 518/781 [00:27<00:13, 19.14it/s]
 67%|######6   | 520/781 [00:27<00:23, 11.11it/s]
 67%|######6   | 522/781 [00:28<00:20, 12.71it/s]
 67%|######7   | 524/781 [00:28<00:18, 14.12it/s]
 67%|######7   | 526/781 [00:28<00:16, 15.31it/s]
 68%|######7   | 528/781 [00:28<00:15, 16.27it/s]
 68%|######7   | 530/781 [00:28<00:14, 17.03it/s]
 68%|######8   | 532/781 [00:28<00:14, 17.60it/s]
 68%|######8   | 534/781 [00:28<00:13, 18.04it/s]
 69%|######8   | 536/781 [00:28<00:13, 18.34it/s]
 69%|######8   | 538/781 [00:28<00:13, 18.57it/s]
 69%|######9   | 540/781 [00:28<00:12, 18.71it/s]
 69%|######9   | 542/781 [00:29<00:12, 18.84it/s]
 70%|######9   | 544/781 [00:29<00:12, 18.93it/s]
 70%|######9   | 546/781 [00:29<00:12, 19.02it/s]
 70%|#######   | 548/781 [00:29<00:12, 19.03it/s]
 70%|#######   | 550/781 [00:29<00:12, 19.07it/s]
 71%|#######   | 552/781 [00:29<00:11, 19.10it/s]
 71%|#######   | 554/781 [00:29<00:11, 19.10it/s]
 71%|#######1  | 556/781 [00:29<00:11, 19.13it/s]
 71%|#######1  | 558/781 [00:29<00:11, 19.15it/s]
 72%|#######1  | 560/781 [00:30<00:11, 19.15it/s]
 72%|#######1  | 562/781 [00:30<00:11, 19.18it/s]
 72%|#######2  | 564/781 [00:30<00:11, 19.20it/s]
 72%|#######2  | 566/781 [00:30<00:11, 19.21it/s]
 73%|#######2  | 568/781 [00:30<00:11, 19.18it/s]
 73%|#######2  | 570/781 [00:30<00:11, 19.18it/s]
 73%|#######3  | 572/781 [00:30<00:10, 19.17it/s]
 73%|#######3  | 574/781 [00:30<00:10, 19.19it/s]
 74%|#######3  | 576/781 [00:30<00:10, 19.15it/s]
 74%|#######4  | 578/781 [00:30<00:10, 19.15it/s]
 74%|#######4  | 580/781 [00:31<00:10, 19.05it/s]
 75%|#######4  | 582/781 [00:31<00:10, 19.06it/s]
 75%|#######4  | 584/781 [00:31<00:10, 18.96it/s]
 75%|#######5  | 586/781 [00:31<00:10, 18.98it/s]
 75%|#######5  | 588/781 [00:31<00:10, 18.95it/s]
 76%|#######5  | 590/781 [00:31<00:10, 18.98it/s]
 76%|#######5  | 592/781 [00:31<00:10, 18.83it/s]
 76%|#######6  | 594/781 [00:31<00:09, 18.92it/s]
 76%|#######6  | 596/781 [00:31<00:09, 18.99it/s]
 77%|#######6  | 598/781 [00:31<00:09, 18.84it/s]
 77%|#######6  | 600/781 [00:32<00:09, 18.80it/s]
 77%|#######7  | 602/781 [00:32<00:09, 18.82it/s]
 77%|#######7  | 604/781 [00:32<00:09, 18.71it/s]
 78%|#######7  | 606/781 [00:32<00:09, 18.57it/s]
 78%|#######7  | 608/781 [00:32<00:09, 18.73it/s]
 78%|#######8  | 610/781 [00:32<00:09, 18.81it/s]
 78%|#######8  | 612/781 [00:32<00:08, 18.90it/s]
 79%|#######8  | 614/781 [00:32<00:08, 18.98it/s]
 79%|#######8  | 616/781 [00:32<00:08, 19.07it/s]
 79%|#######9  | 618/781 [00:33<00:08, 19.08it/s]
 79%|#######9  | 620/781 [00:33<00:08, 19.10it/s]
 80%|#######9  | 622/781 [00:33<00:08, 19.12it/s]
 80%|#######9  | 624/781 [00:33<00:08, 19.07it/s]
 80%|########  | 626/781 [00:33<00:08, 18.95it/s]
 80%|########  | 628/781 [00:33<00:08, 18.89it/s]
 81%|########  | 630/781 [00:33<00:08, 18.83it/s]
 81%|########  | 632/781 [00:33<00:07, 18.85it/s]
 81%|########1 | 634/781 [00:33<00:07, 18.81it/s]
 81%|########1 | 636/781 [00:34<00:07, 18.77it/s]
 82%|########1 | 638/781 [00:34<00:07, 18.75it/s]
 82%|########1 | 640/781 [00:34<00:07, 18.71it/s]
 82%|########2 | 642/781 [00:34<00:07, 18.72it/s]
 82%|########2 | 644/781 [00:34<00:07, 18.71it/s]
 83%|########2 | 646/781 [00:34<00:07, 18.72it/s]
 83%|########2 | 648/781 [00:34<00:07, 18.73it/s]
 83%|########3 | 650/781 [00:34<00:07, 18.71it/s]
 83%|########3 | 652/781 [00:34<00:06, 18.75it/s]
 84%|########3 | 654/781 [00:34<00:06, 18.87it/s]
 84%|########3 | 656/781 [00:35<00:06, 18.97it/s]
 84%|########4 | 658/781 [00:35<00:06, 19.04it/s]
 85%|########4 | 660/781 [00:35<00:06, 19.11it/s]
 85%|########4 | 662/781 [00:35<00:06, 19.16it/s]
 85%|########5 | 664/781 [00:35<00:06, 19.18it/s]
 85%|########5 | 666/781 [00:35<00:06, 19.07it/s]
 86%|########5 | 668/781 [00:35<00:05, 19.09it/s]
 86%|########5 | 670/781 [00:35<00:05, 19.11it/s]
 86%|########6 | 672/781 [00:35<00:05, 19.14it/s]
 86%|########6 | 674/781 [00:36<00:05, 18.84it/s]
 87%|########6 | 676/781 [00:36<00:05, 18.77it/s]
 87%|########6 | 678/781 [00:36<00:05, 18.81it/s]
 87%|########7 | 680/781 [00:36<00:05, 18.94it/s]
 87%|########7 | 682/781 [00:36<00:05, 18.92it/s]
 88%|########7 | 684/781 [00:36<00:05, 18.85it/s]
 88%|########7 | 686/781 [00:36<00:05, 18.79it/s]
 88%|########8 | 688/781 [00:36<00:04, 18.75it/s]
 88%|########8 | 690/781 [00:36<00:04, 18.73it/s]
 89%|########8 | 692/781 [00:36<00:04, 18.77it/s]
 89%|########8 | 694/781 [00:37<00:04, 18.86it/s]
 89%|########9 | 696/781 [00:37<00:04, 18.93it/s]
 89%|########9 | 698/781 [00:37<00:04, 18.96it/s]
 90%|########9 | 700/781 [00:37<00:04, 19.00it/s]
 90%|########9 | 702/781 [00:37<00:04, 19.02it/s]
 90%|######### | 704/781 [00:37<00:04, 19.07it/s]
 90%|######### | 706/781 [00:37<00:03, 19.06it/s]
 91%|######### | 708/781 [00:37<00:03, 19.08it/s]
 91%|######### | 710/781 [00:37<00:03, 19.10it/s]
 91%|#########1| 712/781 [00:38<00:03, 19.11it/s]
 91%|#########1| 714/781 [00:38<00:03, 19.09it/s]
 92%|#########1| 716/781 [00:38<00:03, 19.10it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.09it/s]
 92%|#########2| 720/781 [00:38<00:03, 19.10it/s]
 92%|#########2| 722/781 [00:38<00:03, 19.08it/s]
 93%|#########2| 724/781 [00:38<00:02, 19.06it/s]
 93%|#########2| 726/781 [00:38<00:02, 19.06it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.05it/s]
 93%|#########3| 730/781 [00:38<00:02, 19.03it/s]
 94%|#########3| 732/781 [00:39<00:02, 19.04it/s]
 94%|#########3| 734/781 [00:39<00:02, 19.05it/s]
 94%|#########4| 736/781 [00:39<00:02, 19.00it/s]
 94%|#########4| 738/781 [00:39<00:02, 18.98it/s]
 95%|#########4| 740/781 [00:39<00:02, 19.01it/s]
 95%|#########5| 742/781 [00:39<00:02, 19.03it/s]
 95%|#########5| 744/781 [00:39<00:01, 18.95it/s]
 96%|#########5| 746/781 [00:39<00:01, 19.00it/s]
 96%|#########5| 748/781 [00:39<00:01, 19.05it/s]
 96%|#########6| 750/781 [00:40<00:01, 19.04it/s]
 96%|#########6| 752/781 [00:40<00:01, 18.91it/s]
 97%|#########6| 754/781 [00:40<00:01, 18.97it/s]
 97%|#########6| 756/781 [00:40<00:01, 19.00it/s]
 97%|#########7| 758/781 [00:40<00:01, 19.05it/s]
 97%|#########7| 760/781 [00:40<00:01, 19.06it/s]
 98%|#########7| 762/781 [00:40<00:00, 19.08it/s]
 98%|#########7| 764/781 [00:40<00:00, 19.11it/s]
 98%|#########8| 766/781 [00:40<00:00, 19.00it/s]
 98%|#########8| 768/781 [00:40<00:00, 18.87it/s]
 99%|#########8| 770/781 [00:41<00:00, 18.95it/s]
 99%|#########8| 772/781 [00:41<00:00, 18.90it/s]
 99%|#########9| 774/781 [00:41<00:00, 18.88it/s]
 99%|#########9| 776/781 [00:41<00:00, 18.88it/s]
100%|#########9| 778/781 [00:41<00:00, 18.95it/s]
100%|#########9| 780/781 [00:41<00:00, 18.84it/s]
100%|##########| 781/781 [00:43<00:00, 17.86it/s]
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<01:02,  1.45s/it]
Validation Progress:   2%|2         | 1/44 [00:01<01:02,  1.46s/it]
Validation Progress:   2%|2         | 1/44 [00:01<01:02,  1.46s/it]
Validation Progress:   2%|2         | 1/44 [00:01<01:02,  1.46s/it]
Validation Progress:   2%|2         | 1/44 [00:01<01:03,  1.47s/it]
Validation Progress:   2%|2         | 1/44 [00:01<01:04,  1.50s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:03,  1.52s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:03,  1.52s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:03,  1.52s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.52s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.52s/it]
Validation Progress:   5%|4         | 2/44 [00:03<01:04,  1.54s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.46s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.46s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.46s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.46s/it]
Validation Progress:   7%|6         | 3/44 [00:04<00:59,  1.46s/it]
Validation Progress:   7%|6         | 3/44 [00:04<01:00,  1.47s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:55,  1.39s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:55,  1.39s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:55,  1.39s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:55,  1.39s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:55,  1.39s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:55,  1.39s/it]
Validation Progress:  11%|#1        | 5/44 [00:07<00:56,  1.44s/it]
Validation Progress:  11%|#1        | 5/44 [00:07<00:56,  1.44s/it]
Validation Progress:  11%|#1        | 5/44 [00:07<00:56,  1.44s/it]
Validation Progress:  11%|#1        | 5/44 [00:07<00:56,  1.44s/it]
Validation Progress:  11%|#1        | 5/44 [00:07<00:56,  1.44s/it]
Validation Progress:  11%|#1        | 5/44 [00:07<00:56,  1.44s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:53,  1.42s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:53,  1.42s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:53,  1.42s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:53,  1.42s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:53,  1.42s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:54,  1.42s/it]
Validation Progress:  16%|#5        | 7/44 [00:10<00:54,  1.47s/it]
Validation Progress:  16%|#5        | 7/44 [00:10<00:54,  1.47s/it]
Validation Progress:  16%|#5        | 7/44 [00:10<00:54,  1.47s/it]
Validation Progress:  16%|#5        | 7/44 [00:10<00:54,  1.47s/it]
Validation Progress:  16%|#5        | 7/44 [00:10<00:54,  1.47s/it]
Validation Progress:  16%|#5        | 7/44 [00:10<00:54,  1.47s/it]
Validation Progress:  18%|#8        | 8/44 [00:11<00:53,  1.50s/it]
Validation Progress:  18%|#8        | 8/44 [00:11<00:53,  1.49s/it]
Validation Progress:  18%|#8        | 8/44 [00:11<00:53,  1.49s/it]
Validation Progress:  18%|#8        | 8/44 [00:11<00:53,  1.49s/it]
Validation Progress:  18%|#8        | 8/44 [00:11<00:53,  1.50s/it]
Validation Progress:  18%|#8        | 8/44 [00:11<00:54,  1.51s/it]
Validation Progress:  20%|##        | 9/44 [00:13<00:50,  1.44s/it]
Validation Progress:  20%|##        | 9/44 [00:13<00:50,  1.44s/it]
Validation Progress:  20%|##        | 9/44 [00:13<00:50,  1.44s/it]
Validation Progress:  20%|##        | 9/44 [00:13<00:50,  1.44s/it]
Validation Progress:  20%|##        | 9/44 [00:13<00:50,  1.44s/it]
Validation Progress:  20%|##        | 9/44 [00:13<00:50,  1.45s/it]
Validation Progress:  23%|##2       | 10/44 [00:14<00:48,  1.42s/it]
Validation Progress:  23%|##2       | 10/44 [00:14<00:48,  1.42s/it]
Validation Progress:  23%|##2       | 10/44 [00:14<00:48,  1.42s/it]
Validation Progress:  23%|##2       | 10/44 [00:14<00:48,  1.42s/it]
Validation Progress:  23%|##2       | 10/44 [00:14<00:48,  1.42s/it]
Validation Progress:  23%|##2       | 10/44 [00:14<00:48,  1.43s/it]
Validation Progress:  25%|##5       | 11/44 [00:15<00:48,  1.46s/it]
Validation Progress:  25%|##5       | 11/44 [00:15<00:47,  1.43s/it]
Validation Progress:  25%|##5       | 11/44 [00:15<00:48,  1.46s/it]
Validation Progress:  25%|##5       | 11/44 [00:15<00:48,  1.46s/it]
Validation Progress:  25%|##5       | 11/44 [00:15<00:48,  1.46s/it]
Validation Progress:  25%|##5       | 11/44 [00:15<00:48,  1.46s/it]
Validation Progress:  27%|##7       | 12/44 [00:17<00:47,  1.49s/it]
Validation Progress:  27%|##7       | 12/44 [00:17<00:47,  1.50s/it]
Validation Progress:  27%|##7       | 12/44 [00:17<00:47,  1.49s/it]
Validation Progress:  27%|##7       | 12/44 [00:17<00:47,  1.49s/it]
Validation Progress:  27%|##7       | 12/44 [00:17<00:47,  1.49s/it]
Validation Progress:  27%|##7       | 12/44 [00:17<00:47,  1.48s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:47,  1.54s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:47,  1.54s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:47,  1.54s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:47,  1.53s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:47,  1.54s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:47,  1.54s/it]
Validation Progress:  32%|###1      | 14/44 [00:20<00:45,  1.52s/it]
Validation Progress:  32%|###1      | 14/44 [00:20<00:45,  1.52s/it]
Validation Progress:  32%|###1      | 14/44 [00:20<00:45,  1.52s/it]
Validation Progress:  32%|###1      | 14/44 [00:20<00:45,  1.52s/it]
Validation Progress:  32%|###1      | 14/44 [00:20<00:45,  1.51s/it]
Validation Progress:  32%|###1      | 14/44 [00:20<00:45,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:43,  1.50s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:43,  1.50s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:43,  1.50s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:43,  1.50s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:43,  1.50s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:43,  1.50s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:42,  1.52s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:42,  1.53s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:42,  1.52s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:42,  1.53s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:42,  1.52s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:42,  1.53s/it]
Validation Progress:  39%|###8      | 17/44 [00:25<00:40,  1.51s/it]
Validation Progress:  39%|###8      | 17/44 [00:25<00:40,  1.52s/it]
Validation Progress:  39%|###8      | 17/44 [00:25<00:40,  1.52s/it]
Validation Progress:  39%|###8      | 17/44 [00:25<00:40,  1.52s/it]
Validation Progress:  39%|###8      | 17/44 [00:25<00:40,  1.51s/it]
Validation Progress:  39%|###8      | 17/44 [00:25<00:40,  1.52s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:38,  1.50s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:38,  1.50s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:38,  1.50s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:38,  1.50s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:38,  1.50s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:38,  1.50s/it]
Validation Progress:  43%|####3     | 19/44 [00:28<00:38,  1.53s/it]
Validation Progress:  43%|####3     | 19/44 [00:28<00:38,  1.53s/it]
Validation Progress:  43%|####3     | 19/44 [00:28<00:38,  1.53s/it]
Validation Progress:  43%|####3     | 19/44 [00:28<00:38,  1.53s/it]
Validation Progress:  43%|####3     | 19/44 [00:28<00:38,  1.53s/it]
Validation Progress:  43%|####3     | 19/44 [00:28<00:38,  1.53s/it]
Validation Progress:  45%|####5     | 20/44 [00:29<00:36,  1.52s/it]
Validation Progress:  45%|####5     | 20/44 [00:29<00:36,  1.52s/it]
Validation Progress:  45%|####5     | 20/44 [00:29<00:36,  1.52s/it]
Validation Progress:  45%|####5     | 20/44 [00:29<00:36,  1.52s/it]
Validation Progress:  45%|####5     | 20/44 [00:29<00:36,  1.52s/it]
Validation Progress:  45%|####5     | 20/44 [00:29<00:36,  1.52s/it]
Validation Progress:  48%|####7     | 21/44 [00:31<00:34,  1.51s/it]
Validation Progress:  48%|####7     | 21/44 [00:31<00:34,  1.51s/it]
Validation Progress:  48%|####7     | 21/44 [00:31<00:34,  1.51s/it]
Validation Progress:  48%|####7     | 21/44 [00:31<00:34,  1.51s/it]
Validation Progress:  48%|####7     | 21/44 [00:31<00:34,  1.51s/it]
Validation Progress:  48%|####7     | 21/44 [00:31<00:34,  1.51s/it]
Validation Progress:  50%|#####     | 22/44 [00:32<00:34,  1.55s/it]
Validation Progress:  50%|#####     | 22/44 [00:32<00:34,  1.55s/it]
Validation Progress:  50%|#####     | 22/44 [00:32<00:34,  1.55s/it]
Validation Progress:  50%|#####     | 22/44 [00:32<00:34,  1.55s/it]
Validation Progress:  50%|#####     | 22/44 [00:32<00:34,  1.55s/it]
Validation Progress:  50%|#####     | 22/44 [00:32<00:34,  1.55s/it]
Validation Progress:  52%|#####2    | 23/44 [00:34<00:32,  1.53s/it]
Validation Progress:  52%|#####2    | 23/44 [00:34<00:32,  1.53s/it]
Validation Progress:  52%|#####2    | 23/44 [00:34<00:32,  1.53s/it]
Validation Progress:  52%|#####2    | 23/44 [00:34<00:32,  1.53s/it]
Validation Progress:  52%|#####2    | 23/44 [00:34<00:32,  1.53s/it]
Validation Progress:  52%|#####2    | 23/44 [00:34<00:32,  1.53s/it]
Validation Progress:  55%|#####4    | 24/44 [00:35<00:30,  1.54s/it]
Validation Progress:  55%|#####4    | 24/44 [00:35<00:30,  1.54s/it]
Validation Progress:  55%|#####4    | 24/44 [00:35<00:30,  1.54s/it]
Validation Progress:  55%|#####4    | 24/44 [00:35<00:30,  1.54s/it]
Validation Progress:  55%|#####4    | 24/44 [00:35<00:30,  1.54s/it]
Validation Progress:  55%|#####4    | 24/44 [00:35<00:30,  1.54s/it]
Validation Progress:  57%|#####6    | 25/44 [00:37<00:28,  1.51s/it]
Validation Progress:  57%|#####6    | 25/44 [00:37<00:28,  1.51s/it]
Validation Progress:  57%|#####6    | 25/44 [00:37<00:28,  1.51s/it]
Validation Progress:  57%|#####6    | 25/44 [00:37<00:28,  1.51s/it]
Validation Progress:  57%|#####6    | 25/44 [00:37<00:28,  1.51s/it]
Validation Progress:  57%|#####6    | 25/44 [00:37<00:28,  1.51s/it]
Validation Progress:  59%|#####9    | 26/44 [00:38<00:26,  1.48s/it]
Validation Progress:  59%|#####9    | 26/44 [00:38<00:26,  1.49s/it]
Validation Progress:  59%|#####9    | 26/44 [00:38<00:26,  1.49s/it]
Validation Progress:  59%|#####9    | 26/44 [00:38<00:26,  1.49s/it]
Validation Progress:  59%|#####9    | 26/44 [00:38<00:26,  1.49s/it]
Validation Progress:  59%|#####9    | 26/44 [00:38<00:26,  1.49s/it]
Validation Progress:  61%|######1   | 27/44 [00:40<00:25,  1.48s/it]
Validation Progress:  61%|######1   | 27/44 [00:40<00:25,  1.48s/it]
Validation Progress:  61%|######1   | 27/44 [00:40<00:25,  1.48s/it]
Validation Progress:  61%|######1   | 27/44 [00:40<00:25,  1.48s/it]
Validation Progress:  61%|######1   | 27/44 [00:40<00:25,  1.48s/it]
Validation Progress:  61%|######1   | 27/44 [00:40<00:25,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [00:41<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [00:41<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [00:41<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [00:41<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [00:41<00:23,  1.48s/it]
Validation Progress:  64%|######3   | 28/44 [00:41<00:23,  1.49s/it]
Validation Progress:  66%|######5   | 29/44 [00:43<00:22,  1.49s/it]
Validation Progress:  66%|######5   | 29/44 [00:43<00:22,  1.49s/it]
Validation Progress:  66%|######5   | 29/44 [00:43<00:22,  1.49s/it]
Validation Progress:  66%|######5   | 29/44 [00:43<00:22,  1.49s/it]
Validation Progress:  66%|######5   | 29/44 [00:43<00:22,  1.49s/it]
Validation Progress:  66%|######5   | 29/44 [00:43<00:22,  1.49s/it]
Validation Progress:  68%|######8   | 30/44 [00:44<00:20,  1.49s/it]
Validation Progress:  68%|######8   | 30/44 [00:44<00:20,  1.49s/it]
Validation Progress:  68%|######8   | 30/44 [00:44<00:20,  1.49s/it]
Validation Progress:  68%|######8   | 30/44 [00:44<00:20,  1.49s/it]
Validation Progress:  68%|######8   | 30/44 [00:44<00:20,  1.49s/it]
Validation Progress:  68%|######8   | 30/44 [00:44<00:20,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [00:46<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [00:46<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [00:46<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [00:46<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [00:46<00:19,  1.49s/it]
Validation Progress:  70%|#######   | 31/44 [00:46<00:19,  1.50s/it]
Validation Progress:  73%|#######2  | 32/44 [00:47<00:17,  1.49s/it]
Validation Progress:  73%|#######2  | 32/44 [00:47<00:17,  1.49s/it]
Validation Progress:  73%|#######2  | 32/44 [00:47<00:17,  1.49s/it]
Validation Progress:  73%|#######2  | 32/44 [00:47<00:17,  1.49s/it]
Validation Progress:  73%|#######2  | 32/44 [00:47<00:17,  1.49s/it]
Validation Progress:  73%|#######2  | 32/44 [00:47<00:17,  1.49s/it]
Validation Progress:  75%|#######5  | 33/44 [00:49<00:16,  1.53s/it]
Validation Progress:  75%|#######5  | 33/44 [00:49<00:16,  1.53s/it]
Validation Progress:  75%|#######5  | 33/44 [00:49<00:16,  1.53s/it]
Validation Progress:  75%|#######5  | 33/44 [00:49<00:16,  1.53s/it]
Validation Progress:  75%|#######5  | 33/44 [00:49<00:16,  1.53s/it]
Validation Progress:  75%|#######5  | 33/44 [00:49<00:16,  1.53s/it]
Validation Progress:  77%|#######7  | 34/44 [00:50<00:15,  1.55s/it]
Validation Progress:  77%|#######7  | 34/44 [00:50<00:15,  1.55s/it]
Validation Progress:  77%|#######7  | 34/44 [00:50<00:15,  1.55s/it]
Validation Progress:  77%|#######7  | 34/44 [00:50<00:15,  1.55s/it]
Validation Progress:  77%|#######7  | 34/44 [00:50<00:15,  1.55s/it]
Validation Progress:  77%|#######7  | 34/44 [00:50<00:15,  1.55s/it]
Validation Progress:  80%|#######9  | 35/44 [00:52<00:14,  1.59s/it]
Validation Progress:  80%|#######9  | 35/44 [00:52<00:14,  1.59s/it]
Validation Progress:  80%|#######9  | 35/44 [00:52<00:14,  1.59s/it]
Validation Progress:  80%|#######9  | 35/44 [00:52<00:14,  1.59s/it]
Validation Progress:  80%|#######9  | 35/44 [00:52<00:14,  1.59s/it]
Validation Progress:  80%|#######9  | 35/44 [00:52<00:14,  1.59s/it]
Validation Progress:  82%|########1 | 36/44 [00:54<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [00:54<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [00:54<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [00:54<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [00:54<00:12,  1.56s/it]
Validation Progress:  82%|########1 | 36/44 [00:54<00:12,  1.56s/it]
Validation Progress:  84%|########4 | 37/44 [00:55<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [00:55<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [00:55<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [00:55<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [00:55<00:10,  1.53s/it]
Validation Progress:  84%|########4 | 37/44 [00:55<00:10,  1.53s/it]
Validation Progress:  86%|########6 | 38/44 [00:57<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [00:57<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [00:57<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [00:57<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [00:57<00:09,  1.50s/it]
Validation Progress:  86%|########6 | 38/44 [00:57<00:09,  1.50s/it]
Validation Progress:  89%|########8 | 39/44 [00:58<00:07,  1.55s/it]
Validation Progress:  89%|########8 | 39/44 [00:58<00:07,  1.55s/it]
Validation Progress:  89%|########8 | 39/44 [00:58<00:07,  1.55s/it]
Validation Progress:  89%|########8 | 39/44 [00:58<00:07,  1.55s/it]
Validation Progress:  89%|########8 | 39/44 [00:58<00:07,  1.55s/it]
Validation Progress:  89%|########8 | 39/44 [00:58<00:07,  1.56s/it]
Validation Progress:  91%|######### | 40/44 [01:00<00:06,  1.57s/it]
Validation Progress:  91%|######### | 40/44 [01:00<00:06,  1.57s/it]
Validation Progress:  91%|######### | 40/44 [01:00<00:06,  1.57s/it]
Validation Progress:  91%|######### | 40/44 [01:00<00:06,  1.57s/it]
Validation Progress:  91%|######### | 40/44 [01:00<00:06,  1.57s/it]
Validation Progress:  91%|######### | 40/44 [01:00<00:06,  1.57s/it]
Validation Progress:  93%|#########3| 41/44 [01:01<00:04,  1.54s/it]
Validation Progress:  93%|#########3| 41/44 [01:01<00:04,  1.54s/it]
Validation Progress:  93%|#########3| 41/44 [01:01<00:04,  1.54s/it]
Validation Progress:  93%|#########3| 41/44 [01:01<00:04,  1.54s/it]
Validation Progress:  93%|#########3| 41/44 [01:01<00:04,  1.54s/it]
Validation Progress:  93%|#########3| 41/44 [01:01<00:04,  1.54s/it]
Validation Progress:  95%|#########5| 42/44 [01:03<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [01:03<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [01:03<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [01:03<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [01:03<00:03,  1.52s/it]
Validation Progress:  95%|#########5| 42/44 [01:03<00:03,  1.52s/it]
Validation Progress:  98%|#########7| 43/44 [01:04<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [01:04<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [01:04<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [01:04<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [01:04<00:01,  1.51s/it]
Validation Progress:  98%|#########7| 43/44 [01:04<00:01,  1.51s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:08<00:00,  2.06s/it]
Validation Progress: 100%|##########| 44/44 [01:08<00:00,  1.55s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:08<00:00,  2.06s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:08<00:00,  2.06s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:08<00:00,  2.06s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:08<00:00,  2.06s/it]
Validation Progress: 100%|##########| 44/44 [01:08<00:00,  1.55s/it]
Validation Progress: 100%|##########| 44/44 [01:08<00:00,  1.55s/it]
Validation Progress: 100%|##########| 44/44 [01:08<00:00,  1.55s/it]
Validation Progress: 100%|##########| 44/44 [01:08<00:00,  1.55s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:08<00:00,  2.06s/it]
Validation Progress: 100%|##########| 44/44 [01:08<00:00,  1.55s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:02<00:00,  2.38s/it]
100%|##########| 1/1 [00:02<00:00,  2.38s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:10, 14.34it/s]
  3%|2         | 4/157 [00:00<00:10, 14.54it/s]
  4%|3         | 6/157 [00:00<00:10, 14.96it/s]
  5%|5         | 8/157 [00:00<00:09, 16.15it/s]
  6%|6         | 10/157 [00:00<00:08, 16.80it/s]
  8%|7         | 12/157 [00:00<00:08, 17.24it/s]
  9%|8         | 14/157 [00:00<00:08, 17.57it/s]
 10%|#         | 16/157 [00:00<00:07, 17.76it/s]
 11%|#1        | 18/157 [00:01<00:07, 17.88it/s]
 13%|#2        | 20/157 [00:01<00:07, 18.07it/s]
 14%|#4        | 22/157 [00:01<00:07, 18.23it/s]
 15%|#5        | 24/157 [00:01<00:07, 18.34it/s]
 17%|#6        | 26/157 [00:01<00:07, 18.41it/s]
 18%|#7        | 28/157 [00:01<00:06, 18.45it/s]
 19%|#9        | 30/157 [00:01<00:06, 18.52it/s]
 20%|##        | 32/157 [00:01<00:06, 18.55it/s]
 22%|##1       | 34/157 [00:01<00:06, 18.58it/s]
 23%|##2       | 36/157 [00:02<00:06, 18.60it/s]
 24%|##4       | 38/157 [00:02<00:06, 18.61it/s]
 25%|##5       | 40/157 [00:02<00:06, 18.58it/s]
 27%|##6       | 42/157 [00:02<00:06, 18.57it/s]
 28%|##8       | 44/157 [00:02<00:06, 18.53it/s]
 29%|##9       | 46/157 [00:02<00:05, 18.55it/s]
 31%|###       | 48/157 [00:02<00:05, 18.56it/s]
 32%|###1      | 50/157 [00:02<00:05, 18.55it/s]
 33%|###3      | 52/157 [00:02<00:05, 18.59it/s]
 34%|###4      | 54/157 [00:03<00:05, 18.60it/s]
 36%|###5      | 56/157 [00:03<00:05, 18.61it/s]
 37%|###6      | 58/157 [00:03<00:05, 18.62it/s]
 38%|###8      | 60/157 [00:03<00:05, 18.64it/s]
 39%|###9      | 62/157 [00:03<00:05, 18.64it/s]
 41%|####      | 64/157 [00:03<00:04, 18.65it/s]
 42%|####2     | 66/157 [00:03<00:04, 18.62it/s]
 43%|####3     | 68/157 [00:03<00:04, 18.62it/s]
 45%|####4     | 70/157 [00:03<00:04, 18.64it/s]
 46%|####5     | 72/157 [00:03<00:04, 18.62it/s]
 47%|####7     | 74/157 [00:04<00:04, 18.61it/s]
 48%|####8     | 76/157 [00:04<00:04, 18.62it/s]
 50%|####9     | 78/157 [00:04<00:04, 18.59it/s]
 51%|#####     | 80/157 [00:04<00:04, 18.56it/s]
 52%|#####2    | 82/157 [00:04<00:04, 18.54it/s]
 54%|#####3    | 84/157 [00:04<00:03, 18.58it/s]
 55%|#####4    | 86/157 [00:04<00:03, 18.60it/s]
 56%|#####6    | 88/157 [00:04<00:03, 18.65it/s]
 57%|#####7    | 90/157 [00:04<00:03, 18.65it/s]
 59%|#####8    | 92/157 [00:05<00:03, 18.64it/s]
 60%|#####9    | 94/157 [00:05<00:03, 18.63it/s]
 61%|######1   | 96/157 [00:05<00:03, 18.60it/s]
 62%|######2   | 98/157 [00:05<00:03, 18.60it/s]
 64%|######3   | 100/157 [00:05<00:03, 18.59it/s]
 65%|######4   | 102/157 [00:05<00:02, 18.63it/s]
 66%|######6   | 104/157 [00:05<00:02, 18.64it/s]
 68%|######7   | 106/157 [00:05<00:02, 18.65it/s]
 69%|######8   | 108/157 [00:05<00:02, 18.65it/s]
 70%|#######   | 110/157 [00:06<00:02, 18.60it/s]
 71%|#######1  | 112/157 [00:06<00:02, 18.51it/s]
 73%|#######2  | 114/157 [00:06<00:02, 18.45it/s]
 74%|#######3  | 116/157 [00:06<00:02, 18.42it/s]
 75%|#######5  | 118/157 [00:06<00:02, 18.41it/s]
 76%|#######6  | 120/157 [00:06<00:02, 18.40it/s]
 78%|#######7  | 122/157 [00:06<00:01, 18.43it/s]
 79%|#######8  | 124/157 [00:06<00:01, 18.49it/s]
 80%|########  | 126/157 [00:06<00:01, 18.45it/s]
 82%|########1 | 128/157 [00:06<00:01, 18.49it/s]
 83%|########2 | 130/157 [00:07<00:01, 18.48it/s]
 84%|########4 | 132/157 [00:07<00:01, 18.48it/s]
 85%|########5 | 134/157 [00:07<00:01, 18.49it/s]
 87%|########6 | 136/157 [00:07<00:01, 18.50it/s]
 88%|########7 | 138/157 [00:07<00:01, 18.48it/s]
 89%|########9 | 140/157 [00:07<00:00, 18.50it/s]
 90%|######### | 142/157 [00:07<00:00, 18.50it/s]
 92%|#########1| 144/157 [00:07<00:00, 18.28it/s]
 93%|#########2| 146/157 [00:07<00:00, 18.31it/s]
 94%|#########4| 148/157 [00:08<00:00, 18.37it/s]
 96%|#########5| 150/157 [00:08<00:00, 18.34it/s]
 97%|#########6| 152/157 [00:08<00:00, 18.37it/s]
 98%|#########8| 154/157 [00:08<00:00, 18.30it/s]
 99%|#########9| 156/157 [00:08<00:00, 18.32it/s]
100%|##########| 157/157 [00:08<00:00, 18.39it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:46, 16.71it/s]
  1%|          | 4/781 [00:00<00:46, 16.78it/s]
  1%|          | 6/781 [00:00<00:46, 16.76it/s]
  1%|1         | 8/781 [00:00<00:45, 17.06it/s]
  1%|1         | 10/781 [00:00<00:45, 16.99it/s]
  2%|1         | 12/781 [00:00<00:45, 17.01it/s]
  2%|1         | 14/781 [00:00<00:45, 16.96it/s]
  2%|2         | 16/781 [00:00<00:44, 17.17it/s]
  2%|2         | 18/781 [00:01<00:44, 17.30it/s]
  3%|2         | 20/781 [00:01<00:44, 17.26it/s]
  3%|2         | 22/781 [00:01<00:43, 17.37it/s]
  3%|3         | 24/781 [00:01<00:43, 17.47it/s]
  3%|3         | 26/781 [00:01<00:42, 17.56it/s]
  4%|3         | 28/781 [00:01<00:42, 17.60it/s]
  4%|3         | 30/781 [00:01<00:42, 17.64it/s]
  4%|4         | 32/781 [00:01<00:42, 17.65it/s]
  4%|4         | 34/781 [00:01<00:42, 17.66it/s]
  5%|4         | 36/781 [00:02<00:42, 17.68it/s]
  5%|4         | 38/781 [00:02<00:41, 17.71it/s]
  5%|5         | 40/781 [00:02<00:41, 17.72it/s]
  5%|5         | 42/781 [00:02<00:41, 17.75it/s]
  6%|5         | 44/781 [00:02<00:41, 17.79it/s]
  6%|5         | 46/781 [00:02<00:41, 17.78it/s]
  6%|6         | 48/781 [00:02<00:41, 17.77it/s]
  6%|6         | 50/781 [00:02<00:41, 17.77it/s]
  7%|6         | 52/781 [00:02<00:41, 17.77it/s]
  7%|6         | 54/781 [00:03<00:40, 17.78it/s]
  7%|7         | 56/781 [00:03<00:40, 17.76it/s]
  7%|7         | 58/781 [00:03<00:40, 17.78it/s]
  8%|7         | 60/781 [00:03<00:40, 17.76it/s]
  8%|7         | 62/781 [00:03<00:40, 17.74it/s]
  8%|8         | 64/781 [00:03<00:40, 17.73it/s]
  8%|8         | 66/781 [00:03<00:40, 17.73it/s]
  9%|8         | 68/781 [00:03<00:40, 17.74it/s]
  9%|8         | 70/781 [00:03<00:40, 17.75it/s]
  9%|9         | 72/781 [00:04<00:39, 17.77it/s]
  9%|9         | 74/781 [00:04<00:39, 17.78it/s]
 10%|9         | 76/781 [00:04<00:39, 17.73it/s]
 10%|9         | 78/781 [00:04<00:39, 17.61it/s]
 10%|#         | 80/781 [00:04<00:39, 17.53it/s]
 10%|#         | 82/781 [00:04<00:39, 17.49it/s]
 11%|#         | 84/781 [00:04<00:39, 17.44it/s]
 11%|#1        | 86/781 [00:04<00:39, 17.53it/s]
 11%|#1        | 88/781 [00:05<00:39, 17.59it/s]
 12%|#1        | 90/781 [00:05<00:39, 17.65it/s]
 12%|#1        | 92/781 [00:05<00:38, 17.68it/s]
 12%|#2        | 94/781 [00:05<00:38, 17.71it/s]
 12%|#2        | 96/781 [00:05<00:38, 17.73it/s]
 13%|#2        | 98/781 [00:05<00:38, 17.74it/s]
 13%|#2        | 100/781 [00:05<00:38, 17.70it/s]
 13%|#3        | 102/781 [00:05<00:38, 17.52it/s]
 13%|#3        | 104/781 [00:05<00:38, 17.47it/s]
 14%|#3        | 106/781 [00:06<00:38, 17.45it/s]
 14%|#3        | 108/781 [00:06<00:38, 17.50it/s]
 14%|#4        | 110/781 [00:06<00:38, 17.57it/s]
 14%|#4        | 112/781 [00:06<00:37, 17.61it/s]
 15%|#4        | 114/781 [00:06<00:37, 17.66it/s]
 15%|#4        | 116/781 [00:06<00:37, 17.68it/s]
 15%|#5        | 118/781 [00:06<00:37, 17.69it/s]
 15%|#5        | 120/781 [00:06<00:37, 17.73it/s]
 16%|#5        | 122/781 [00:06<00:37, 17.73it/s]
 16%|#5        | 124/781 [00:07<00:37, 17.73it/s]
 16%|#6        | 126/781 [00:07<00:36, 17.77it/s]
 16%|#6        | 128/781 [00:07<00:36, 17.74it/s]
 17%|#6        | 130/781 [00:07<00:36, 17.74it/s]
 17%|#6        | 132/781 [00:07<00:36, 17.74it/s]
 17%|#7        | 134/781 [00:07<00:36, 17.62it/s]
 17%|#7        | 136/781 [00:07<00:36, 17.53it/s]
 18%|#7        | 138/781 [00:07<00:36, 17.49it/s]
 18%|#7        | 140/781 [00:07<00:36, 17.56it/s]
 18%|#8        | 142/781 [00:08<00:36, 17.63it/s]
 18%|#8        | 144/781 [00:08<00:36, 17.66it/s]
 19%|#8        | 146/781 [00:08<00:35, 17.67it/s]
 19%|#8        | 148/781 [00:08<00:35, 17.71it/s]
 19%|#9        | 150/781 [00:08<00:35, 17.71it/s]
 19%|#9        | 152/781 [00:08<00:35, 17.71it/s]
 20%|#9        | 154/781 [00:08<00:35, 17.72it/s]
 20%|#9        | 156/781 [00:08<00:35, 17.73it/s]
 20%|##        | 158/781 [00:08<00:35, 17.74it/s]
 20%|##        | 160/781 [00:09<00:34, 17.75it/s]
 21%|##        | 162/781 [00:09<00:34, 17.77it/s]
 21%|##        | 164/781 [00:09<00:34, 17.78it/s]
 21%|##1       | 166/781 [00:09<00:34, 17.59it/s]
 22%|##1       | 168/781 [00:09<00:34, 17.61it/s]
 22%|##1       | 170/781 [00:09<00:34, 17.51it/s]
 22%|##2       | 172/781 [00:09<00:35, 17.28it/s]
 22%|##2       | 174/781 [00:09<00:35, 17.29it/s]
 23%|##2       | 176/781 [00:10<00:34, 17.34it/s]
 23%|##2       | 178/781 [00:10<00:34, 17.34it/s]
 23%|##3       | 180/781 [00:10<00:34, 17.44it/s]
 23%|##3       | 182/781 [00:10<00:34, 17.46it/s]
 24%|##3       | 184/781 [00:10<00:34, 17.35it/s]
 24%|##3       | 186/781 [00:10<00:34, 17.31it/s]
 24%|##4       | 188/781 [00:10<00:34, 17.00it/s]
 24%|##4       | 190/781 [00:10<00:34, 17.09it/s]
 25%|##4       | 192/781 [00:10<00:34, 17.16it/s]
 25%|##4       | 194/781 [00:11<00:34, 17.24it/s]
 25%|##5       | 196/781 [00:11<00:33, 17.36it/s]
 25%|##5       | 198/781 [00:11<00:33, 17.40it/s]
 26%|##5       | 200/781 [00:11<00:33, 17.43it/s]
 26%|##5       | 202/781 [00:11<00:33, 17.10it/s]
 26%|##6       | 204/781 [00:11<00:33, 17.02it/s]
 26%|##6       | 206/781 [00:11<00:34, 16.67it/s]
 27%|##6       | 208/781 [00:11<00:34, 16.69it/s]
 27%|##6       | 210/781 [00:11<00:33, 16.97it/s]
 27%|##7       | 212/781 [00:12<00:33, 17.16it/s]
 27%|##7       | 214/781 [00:12<00:33, 17.04it/s]
 28%|##7       | 216/781 [00:12<00:32, 17.21it/s]
 28%|##7       | 218/781 [00:12<00:32, 17.10it/s]
 28%|##8       | 220/781 [00:12<00:32, 17.04it/s]
 28%|##8       | 222/781 [00:12<00:32, 16.96it/s]
 29%|##8       | 224/781 [00:12<00:32, 17.16it/s]
 29%|##8       | 226/781 [00:12<00:32, 17.05it/s]
 29%|##9       | 228/781 [00:13<00:32, 16.99it/s]
 29%|##9       | 230/781 [00:13<00:32, 16.89it/s]
 30%|##9       | 232/781 [00:13<00:32, 17.11it/s]
 30%|##9       | 234/781 [00:13<00:31, 17.26it/s]
 30%|###       | 236/781 [00:13<00:31, 17.38it/s]
 30%|###       | 238/781 [00:13<00:31, 16.97it/s]
 31%|###       | 240/781 [00:13<00:32, 16.90it/s]
 31%|###       | 242/781 [00:13<00:31, 16.89it/s]
 31%|###1      | 244/781 [00:13<00:31, 16.86it/s]
 31%|###1      | 246/781 [00:14<00:32, 16.62it/s]
 32%|###1      | 248/781 [00:14<00:32, 16.53it/s]
 32%|###2      | 250/781 [00:14<00:32, 16.25it/s]
 32%|###2      | 252/781 [00:14<00:32, 16.38it/s]
 33%|###2      | 254/781 [00:14<00:31, 16.76it/s]
 33%|###2      | 256/781 [00:14<00:31, 16.80it/s]
 33%|###3      | 258/781 [00:14<00:30, 17.08it/s]
 33%|###3      | 260/781 [00:14<00:30, 17.18it/s]
 34%|###3      | 262/781 [00:15<00:31, 16.71it/s]
 34%|###3      | 264/781 [00:15<00:30, 16.82it/s]
 34%|###4      | 266/781 [00:15<00:30, 16.71it/s]
 34%|###4      | 268/781 [00:15<00:30, 16.62it/s]
 35%|###4      | 270/781 [00:15<00:30, 16.76it/s]
 35%|###4      | 272/781 [00:15<00:30, 16.88it/s]
 35%|###5      | 274/781 [00:15<00:29, 17.08it/s]
 35%|###5      | 276/781 [00:15<00:29, 16.91it/s]
 36%|###5      | 278/781 [00:16<00:29, 17.10it/s]
 36%|###5      | 280/781 [00:16<00:29, 16.75it/s]
 36%|###6      | 282/781 [00:16<00:29, 16.74it/s]
 36%|###6      | 284/781 [00:16<00:29, 16.76it/s]
 37%|###6      | 286/781 [00:16<00:29, 16.80it/s]
 37%|###6      | 288/781 [00:16<00:29, 16.79it/s]
 37%|###7      | 290/781 [00:16<00:29, 16.81it/s]
 37%|###7      | 292/781 [00:16<00:29, 16.83it/s]
 38%|###7      | 294/781 [00:16<00:29, 16.64it/s]
 38%|###7      | 296/781 [00:17<00:28, 16.94it/s]
 38%|###8      | 298/781 [00:17<00:28, 17.14it/s]
 38%|###8      | 300/781 [00:17<00:28, 16.86it/s]
 39%|###8      | 302/781 [00:17<00:27, 17.12it/s]
 39%|###8      | 304/781 [00:17<00:28, 16.75it/s]
 39%|###9      | 306/781 [00:17<00:28, 16.51it/s]
 39%|###9      | 308/781 [00:17<00:28, 16.80it/s]
 40%|###9      | 310/781 [00:17<00:28, 16.81it/s]
 40%|###9      | 312/781 [00:18<00:27, 16.82it/s]
 40%|####      | 314/781 [00:18<00:27, 17.09it/s]
 40%|####      | 316/781 [00:18<00:27, 17.00it/s]
 41%|####      | 318/781 [00:18<00:27, 16.73it/s]
 41%|####      | 320/781 [00:18<00:27, 16.76it/s]
 41%|####1     | 322/781 [00:18<00:27, 16.78it/s]
 41%|####1     | 324/781 [00:18<00:27, 16.54it/s]
 42%|####1     | 326/781 [00:18<00:27, 16.65it/s]
 42%|####1     | 328/781 [00:18<00:27, 16.74it/s]
 42%|####2     | 330/781 [00:19<00:26, 16.79it/s]
 43%|####2     | 332/781 [00:19<00:26, 17.06it/s]
 43%|####2     | 334/781 [00:19<00:26, 17.00it/s]
 43%|####3     | 336/781 [00:19<00:26, 16.91it/s]
 43%|####3     | 338/781 [00:19<00:26, 17.02it/s]
 44%|####3     | 340/781 [00:19<00:26, 16.94it/s]
 44%|####3     | 342/781 [00:19<00:26, 16.61it/s]
 44%|####4     | 344/781 [00:19<00:26, 16.53it/s]
 44%|####4     | 346/781 [00:20<00:26, 16.49it/s]
 45%|####4     | 348/781 [00:20<00:26, 16.32it/s]
 45%|####4     | 350/781 [00:20<00:26, 16.23it/s]
 45%|####5     | 352/781 [00:20<00:26, 16.20it/s]
 45%|####5     | 354/781 [00:20<00:26, 16.09it/s]
 46%|####5     | 356/781 [00:20<00:26, 16.23it/s]
 46%|####5     | 358/781 [00:20<00:26, 16.16it/s]
 46%|####6     | 360/781 [00:20<00:25, 16.52it/s]
 46%|####6     | 362/781 [00:21<00:25, 16.32it/s]
 47%|####6     | 364/781 [00:21<00:24, 16.68it/s]
 47%|####6     | 366/781 [00:21<00:24, 16.66it/s]
 47%|####7     | 368/781 [00:21<00:25, 16.44it/s]
 47%|####7     | 370/781 [00:21<00:25, 16.34it/s]
 48%|####7     | 372/781 [00:21<00:25, 16.30it/s]
 48%|####7     | 374/781 [00:21<00:25, 16.03it/s]
 48%|####8     | 376/781 [00:21<00:24, 16.26it/s]
 48%|####8     | 378/781 [00:22<00:24, 16.63it/s]
 49%|####8     | 380/781 [00:22<00:24, 16.71it/s]
 49%|####8     | 382/781 [00:22<00:24, 16.53it/s]
 49%|####9     | 384/781 [00:22<00:23, 16.84it/s]
 49%|####9     | 386/781 [00:22<00:23, 16.66it/s]
 50%|####9     | 388/781 [00:22<00:23, 16.53it/s]
 50%|####9     | 390/781 [00:22<00:23, 16.63it/s]
 50%|#####     | 392/781 [00:22<00:23, 16.49it/s]
 50%|#####     | 394/781 [00:22<00:23, 16.44it/s]
 51%|#####     | 396/781 [00:23<00:23, 16.59it/s]
 51%|#####     | 398/781 [00:23<00:22, 16.71it/s]
 51%|#####1    | 400/781 [00:23<00:23, 16.56it/s]
 51%|#####1    | 402/781 [00:23<00:22, 16.89it/s]
 52%|#####1    | 404/781 [00:23<00:22, 17.03it/s]
 52%|#####1    | 406/781 [00:23<00:22, 16.65it/s]
 52%|#####2    | 408/781 [00:23<00:22, 16.58it/s]
 52%|#####2    | 410/781 [00:23<00:22, 16.37it/s]
 53%|#####2    | 412/781 [00:24<00:22, 16.20it/s]
 53%|#####3    | 414/781 [00:24<00:22, 16.29it/s]
 53%|#####3    | 416/781 [00:24<00:22, 16.16it/s]
 54%|#####3    | 418/781 [00:24<00:22, 16.07it/s]
 54%|#####3    | 420/781 [00:24<00:22, 16.00it/s]
 54%|#####4    | 422/781 [00:24<00:21, 16.41it/s]
 54%|#####4    | 424/781 [00:24<00:21, 16.31it/s]
 55%|#####4    | 426/781 [00:24<00:21, 16.28it/s]
 55%|#####4    | 428/781 [00:25<00:21, 16.14it/s]
 55%|#####5    | 430/781 [00:25<00:21, 16.05it/s]
 55%|#####5    | 432/781 [00:25<00:21, 16.17it/s]
 56%|#####5    | 434/781 [00:25<00:21, 16.08it/s]
 56%|#####5    | 436/781 [00:25<00:21, 16.04it/s]
 56%|#####6    | 438/781 [00:25<00:21, 16.16it/s]
 56%|#####6    | 440/781 [00:25<00:20, 16.28it/s]
 57%|#####6    | 442/781 [00:25<00:20, 16.21it/s]
 57%|#####6    | 444/781 [00:26<00:20, 16.43it/s]
 57%|#####7    | 446/781 [00:26<00:20, 16.23it/s]
 57%|#####7    | 448/781 [00:26<00:20, 16.55it/s]
 58%|#####7    | 450/781 [00:26<00:20, 16.36it/s]
 58%|#####7    | 452/781 [00:26<00:20, 16.21it/s]
 58%|#####8    | 454/781 [00:26<00:20, 16.08it/s]
 58%|#####8    | 456/781 [00:26<00:20, 16.05it/s]
 59%|#####8    | 458/781 [00:26<00:20, 16.00it/s]
 59%|#####8    | 460/781 [00:27<00:19, 16.08it/s]
 59%|#####9    | 462/781 [00:27<00:19, 16.14it/s]
 59%|#####9    | 464/781 [00:27<00:19, 16.39it/s]
 60%|#####9    | 466/781 [00:27<00:19, 16.37it/s]
 60%|#####9    | 468/781 [00:27<00:19, 16.35it/s]
 60%|######    | 470/781 [00:27<00:19, 16.27it/s]
 60%|######    | 472/781 [00:27<00:18, 16.46it/s]
 61%|######    | 474/781 [00:27<00:18, 16.39it/s]
 61%|######    | 476/781 [00:28<00:18, 16.56it/s]
 61%|######1   | 478/781 [00:28<00:18, 16.66it/s]
 61%|######1   | 480/781 [00:28<00:17, 16.76it/s]
 62%|######1   | 482/781 [00:28<00:18, 16.49it/s]
 62%|######1   | 484/781 [00:28<00:18, 16.28it/s]
 62%|######2   | 486/781 [00:28<00:18, 16.18it/s]
 62%|######2   | 488/781 [00:28<00:17, 16.34it/s]
 63%|######2   | 490/781 [00:28<00:17, 16.32it/s]
 63%|######2   | 492/781 [00:28<00:17, 16.51it/s]
 63%|######3   | 494/781 [00:29<00:17, 16.66it/s]
 64%|######3   | 496/781 [00:29<00:17, 16.51it/s]
 64%|######3   | 498/781 [00:29<00:17, 16.40it/s]
 64%|######4   | 500/781 [00:29<00:16, 16.57it/s]
 64%|######4   | 502/781 [00:29<00:17, 16.33it/s]
 65%|######4   | 504/781 [00:29<00:17, 16.22it/s]
 65%|######4   | 506/781 [00:29<00:17, 16.04it/s]
 65%|######5   | 508/781 [00:29<00:17, 16.05it/s]
 65%|######5   | 510/781 [00:30<00:16, 16.25it/s]
 66%|######5   | 512/781 [00:30<00:16, 16.27it/s]
 66%|######5   | 514/781 [00:30<00:16, 16.11it/s]
 66%|######6   | 516/781 [00:30<00:16, 16.09it/s]
 66%|######6   | 518/781 [00:30<00:16, 16.04it/s]
 67%|######6   | 520/781 [00:30<00:16, 16.05it/s]
 67%|######6   | 522/781 [00:30<00:16, 15.98it/s]
 67%|######7   | 524/781 [00:30<00:16, 16.04it/s]
 67%|######7   | 526/781 [00:31<00:15, 16.31it/s]
 68%|######7   | 528/781 [00:31<00:15, 16.41it/s]
 68%|######7   | 530/781 [00:31<00:15, 16.37it/s]
 68%|######8   | 532/781 [00:31<00:15, 16.55it/s]
 68%|######8   | 534/781 [00:31<00:14, 16.47it/s]
 69%|######8   | 536/781 [00:31<00:15, 16.31it/s]
 69%|######8   | 538/781 [00:31<00:14, 16.46it/s]
 69%|######9   | 540/781 [00:31<00:14, 16.60it/s]
 69%|######9   | 542/781 [00:32<00:14, 16.71it/s]
 70%|######9   | 544/781 [00:32<00:14, 16.56it/s]
 70%|######9   | 546/781 [00:32<00:14, 16.69it/s]
 70%|#######   | 548/781 [00:32<00:14, 16.24it/s]
 70%|#######   | 550/781 [00:32<00:14, 16.24it/s]
 71%|#######   | 552/781 [00:32<00:13, 16.45it/s]
 71%|#######   | 554/781 [00:32<00:13, 16.81it/s]
 71%|#######1  | 556/781 [00:32<00:13, 16.63it/s]
 71%|#######1  | 558/781 [00:33<00:13, 16.71it/s]
 72%|#######1  | 560/781 [00:33<00:13, 16.57it/s]
 72%|#######1  | 562/781 [00:33<00:13, 16.69it/s]
 72%|#######2  | 564/781 [00:33<00:13, 16.69it/s]
 72%|#######2  | 566/781 [00:33<00:12, 16.67it/s]
 73%|#######2  | 568/781 [00:33<00:12, 16.64it/s]
 73%|#######2  | 570/781 [00:33<00:12, 16.86it/s]
 73%|#######3  | 572/781 [00:33<00:12, 16.84it/s]
 73%|#######3  | 574/781 [00:33<00:12, 16.76it/s]
 74%|#######3  | 576/781 [00:34<00:12, 16.50it/s]
 74%|#######4  | 578/781 [00:34<00:12, 16.32it/s]
 74%|#######4  | 580/781 [00:34<00:12, 16.26it/s]
 75%|#######4  | 582/781 [00:34<00:12, 16.28it/s]
 75%|#######4  | 584/781 [00:34<00:12, 16.29it/s]
 75%|#######5  | 586/781 [00:34<00:11, 16.47it/s]
 75%|#######5  | 588/781 [00:34<00:11, 16.86it/s]
 76%|#######5  | 590/781 [00:34<00:11, 16.94it/s]
 76%|#######5  | 592/781 [00:35<00:11, 16.75it/s]
 76%|#######6  | 594/781 [00:35<00:11, 16.83it/s]
 76%|#######6  | 596/781 [00:35<00:11, 16.55it/s]
 77%|#######6  | 598/781 [00:35<00:11, 16.33it/s]
 77%|#######6  | 600/781 [00:35<00:11, 16.21it/s]
 77%|#######7  | 602/781 [00:35<00:10, 16.32it/s]
 77%|#######7  | 604/781 [00:35<00:10, 16.45it/s]
 78%|#######7  | 606/781 [00:35<00:10, 16.37it/s]
 78%|#######7  | 608/781 [00:36<00:10, 16.54it/s]
 78%|#######8  | 610/781 [00:36<00:10, 16.48it/s]
 78%|#######8  | 612/781 [00:36<00:10, 16.41it/s]
 79%|#######8  | 614/781 [00:36<00:10, 16.38it/s]
 79%|#######8  | 616/781 [00:36<00:09, 16.56it/s]
 79%|#######9  | 618/781 [00:36<00:09, 16.44it/s]
 79%|#######9  | 620/781 [00:36<00:09, 16.64it/s]
 80%|#######9  | 622/781 [00:36<00:09, 16.75it/s]
 80%|#######9  | 624/781 [00:37<00:10, 14.69it/s]
 80%|########  | 626/781 [00:37<00:10, 15.11it/s]
 80%|########  | 628/781 [00:37<00:09, 15.57it/s]
 81%|########  | 630/781 [00:37<00:09, 15.79it/s]
 81%|########  | 632/781 [00:37<00:09, 16.34it/s]
 81%|########1 | 634/781 [00:37<00:08, 16.43it/s]
 81%|########1 | 636/781 [00:37<00:08, 16.34it/s]
 82%|########1 | 638/781 [00:37<00:08, 16.29it/s]
 82%|########1 | 640/781 [00:38<00:08, 16.47it/s]
 82%|########2 | 642/781 [00:38<00:08, 16.62it/s]
 82%|########2 | 644/781 [00:38<00:08, 16.54it/s]
 83%|########2 | 646/781 [00:38<00:08, 16.45it/s]
 83%|########2 | 648/781 [00:38<00:08, 16.38it/s]
 83%|########3 | 650/781 [00:38<00:07, 16.56it/s]
 83%|########3 | 652/781 [00:38<00:07, 16.48it/s]
 84%|########3 | 654/781 [00:38<00:07, 16.31it/s]
 84%|########3 | 656/781 [00:38<00:07, 16.31it/s]
 84%|########4 | 658/781 [00:39<00:07, 16.11it/s]
 85%|########4 | 660/781 [00:39<00:07, 16.17it/s]
 85%|########4 | 662/781 [00:39<00:07, 16.05it/s]
 85%|########5 | 664/781 [00:39<00:07, 16.20it/s]
 85%|########5 | 666/781 [00:39<00:07, 16.12it/s]
 86%|########5 | 668/781 [00:39<00:07, 16.08it/s]
 86%|########5 | 670/781 [00:39<00:06, 16.03it/s]
 86%|########6 | 672/781 [00:39<00:06, 16.04it/s]
 86%|########6 | 674/781 [00:40<00:06, 15.89it/s]
 87%|########6 | 676/781 [00:40<00:06, 15.88it/s]
 87%|########6 | 678/781 [00:40<00:06, 15.84it/s]
 87%|########7 | 680/781 [00:40<00:06, 15.76it/s]
 87%|########7 | 682/781 [00:40<00:06, 15.76it/s]
 88%|########7 | 684/781 [00:40<00:06, 15.93it/s]
 88%|########7 | 686/781 [00:40<00:06, 15.75it/s]
 88%|########8 | 688/781 [00:41<00:05, 16.04it/s]
 88%|########8 | 690/781 [00:41<00:05, 16.04it/s]
 89%|########8 | 692/781 [00:41<00:05, 15.99it/s]
 89%|########8 | 694/781 [00:41<00:05, 16.28it/s]
 89%|########9 | 696/781 [00:41<00:05, 16.24it/s]
 89%|########9 | 698/781 [00:41<00:05, 16.24it/s]
 90%|########9 | 700/781 [00:41<00:04, 16.24it/s]
 90%|########9 | 702/781 [00:41<00:04, 16.26it/s]
 90%|######### | 704/781 [00:41<00:04, 16.29it/s]
 90%|######### | 706/781 [00:42<00:04, 16.30it/s]
 91%|######### | 708/781 [00:42<00:04, 16.54it/s]
 91%|######### | 710/781 [00:42<00:04, 16.66it/s]
 91%|#########1| 712/781 [00:42<00:04, 16.55it/s]
 91%|#########1| 714/781 [00:42<00:04, 16.70it/s]
 92%|#########1| 716/781 [00:42<00:03, 16.56it/s]
 92%|#########1| 718/781 [00:42<00:03, 16.37it/s]
 92%|#########2| 720/781 [00:42<00:03, 16.42it/s]
 92%|#########2| 722/781 [00:43<00:03, 16.51it/s]
 93%|#########2| 724/781 [00:43<00:03, 16.33it/s]
 93%|#########2| 726/781 [00:43<00:03, 16.17it/s]
 93%|#########3| 728/781 [00:43<00:03, 16.30it/s]
 93%|#########3| 730/781 [00:43<00:03, 16.19it/s]
 94%|#########3| 732/781 [00:43<00:02, 16.44it/s]
 94%|#########3| 734/781 [00:43<00:02, 16.24it/s]
 94%|#########4| 736/781 [00:43<00:02, 16.36it/s]
 94%|#########4| 738/781 [00:44<00:02, 16.46it/s]
 95%|#########4| 740/781 [00:44<00:02, 16.28it/s]
 95%|#########5| 742/781 [00:44<00:02, 16.58it/s]
 95%|#########5| 744/781 [00:44<00:02, 16.34it/s]
 96%|#########5| 746/781 [00:44<00:02, 16.41it/s]
 96%|#########5| 748/781 [00:44<00:02, 16.46it/s]
 96%|#########6| 750/781 [00:44<00:01, 16.27it/s]
 96%|#########6| 752/781 [00:44<00:01, 16.13it/s]
 97%|#########6| 754/781 [00:45<00:01, 16.27it/s]
 97%|#########6| 756/781 [00:45<00:01, 16.20it/s]
 97%|#########7| 758/781 [00:45<00:01, 16.31it/s]
 97%|#########7| 760/781 [00:45<00:01, 16.19it/s]
 98%|#########7| 762/781 [00:45<00:01, 16.18it/s]
 98%|#########7| 764/781 [00:45<00:01, 16.06it/s]
 98%|#########8| 766/781 [00:45<00:00, 16.14it/s]
 98%|#########8| 768/781 [00:45<00:00, 16.01it/s]
 99%|#########8| 770/781 [00:46<00:00, 16.19it/s]
 99%|#########8| 772/781 [00:46<00:00, 16.20it/s]
 99%|#########9| 774/781 [00:46<00:00, 16.42it/s]
 99%|#########9| 776/781 [00:46<00:00, 16.59it/s]
100%|#########9| 778/781 [00:46<00:00, 16.50it/s]
100%|#########9| 780/781 [00:46<00:00, 16.39it/s]
100%|##########| 781/781 [00:46<00:00, 16.73it/s]






=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================







=======================================================   RESTART [11-25 01:24:25]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241124_092747-vnmtzxkp
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run gallant-pine-544
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/vnmtzxkp







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================







=======================================================   RESTART [11-25 02:36:22]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241124_103944-8mjgrch3
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run ethereal-hill-545
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/8mjgrch3
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_7 0.0246 ms 100.0%
  triton_mm_8 0.0248 ms 99.1%
  triton_mm_12 0.0248 ms 99.1%
  triton_mm_9 0.0249 ms 98.7%
  triton_mm_11 0.0258 ms 95.2%
  triton_mm_6 0.0259 ms 94.9%
  triton_mm_4 0.0267 ms 92.2%
  triton_mm_3 0.0273 ms 90.0%
  triton_mm_14 0.0273 ms 89.9%
  triton_mm_2 0.0274 ms 89.7%
SingleProcess AUTOTUNE benchmarking takes 2.1911 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2898 1.8470 ms 100.0%
  triton_mm_2897 1.8874 ms 97.9%
  bias_addmm 1.9864 ms 93.0%
  triton_mm_2899 2.0303 ms 91.0%
  triton_mm_2892 2.3001 ms 80.3%
  triton_mm_2894 2.3602 ms 78.3%
  triton_mm_2890 2.4181 ms 76.4%
  addmm 2.4636 ms 75.0%
  triton_mm_2893 2.4717 ms 74.7%
  triton_mm_2895 2.6956 ms 68.5%
SingleProcess AUTOTUNE benchmarking takes 3.2160 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_9 0.0248 ms 100.0%
  triton_mm_11 0.0248 ms 100.0%
  triton_mm_12 0.0249 ms 99.5%
  triton_mm_8 0.0250 ms 99.2%
  triton_mm_7 0.0253 ms 98.0%
  triton_mm_6 0.0260 ms 95.6%
  triton_mm_2 0.0262 ms 94.7%
  triton_mm_4 0.0265 ms 93.6%
  triton_mm_3 0.0272 ms 91.2%
  triton_mm_15 0.0273 ms 90.9%
SingleProcess AUTOTUNE benchmarking takes 2.1746 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2872 0.0232 ms 100.0%
  triton_mm_2880 0.0236 ms 98.2%
  triton_mm_2868 0.0242 ms 95.8%
  triton_mm_2876 0.0245 ms 94.6%
  bias_addmm 0.0277 ms 83.7%
  triton_mm_2871 0.0285 ms 81.3%
  triton_mm_2865 0.0292 ms 79.4%
  addmm 0.0299 ms 77.7%
  triton_mm_2875 0.0299 ms 77.6%
  triton_mm_2867 0.0301 ms 77.2%
SingleProcess AUTOTUNE benchmarking takes 2.3476 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  bias_addmm 1.8195 ms 100.0%
  triton_mm_2898 1.8387 ms 99.0%
  triton_mm_2897 1.8888 ms 96.3%
  triton_mm_2899 2.0363 ms 89.4%
  addmm 2.1937 ms 82.9%
  triton_mm_2892 2.2729 ms 80.0%
  triton_mm_2894 2.3366 ms 77.9%
  triton_mm_2890 2.3990 ms 75.8%
  triton_mm_2893 2.4681 ms 73.7%
  triton_mm_2895 2.6697 ms 68.2%
SingleProcess AUTOTUNE benchmarking takes 3.1787 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0227 ms 100.0%
  triton_mm_2872 0.0228 ms 99.5%
  triton_mm_2868 0.0234 ms 96.7%
  triton_mm_2880 0.0238 ms 95.3%
  bias_addmm 0.0263 ms 86.2%
  triton_mm_2865 0.0270 ms 84.0%
  triton_mm_2871 0.0274 ms 82.6%
  triton_mm_2867 0.0275 ms 82.3%
  triton_mm_2875 0.0276 ms 82.2%
  triton_mm_2866 0.0279 ms 81.3%
SingleProcess AUTOTUNE benchmarking takes 2.3013 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  bias_addmm 1.8150 ms 100.0%
  triton_mm_2898 1.8486 ms 98.2%
  triton_mm_2897 1.8711 ms 97.0%
  triton_mm_2899 2.0569 ms 88.2%
  addmm 2.1838 ms 83.1%
  triton_mm_2892 2.2623 ms 80.2%
  triton_mm_2894 2.3344 ms 77.7%
  triton_mm_2890 2.4098 ms 75.3%
  triton_mm_2893 2.4659 ms 73.6%
  triton_mm_2895 2.6709 ms 68.0%
SingleProcess AUTOTUNE benchmarking takes 3.1754 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2876 0.0225 ms 100.0%
  triton_mm_2872 0.0228 ms 98.9%
  triton_mm_2868 0.0235 ms 96.0%
  triton_mm_2880 0.0235 ms 95.8%
  bias_addmm 0.0261 ms 86.2%
  triton_mm_2871 0.0262 ms 86.1%
  triton_mm_2867 0.0273 ms 82.4%
  triton_mm_2875 0.0273 ms 82.4%
  triton_mm_2865 0.0276 ms 81.6%
  triton_mm_2866 0.0285 ms 79.0%
SingleProcess AUTOTUNE benchmarking takes 2.2990 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(12x3840, 3840x1920)
  triton_mm_2946 0.0233 ms 100.0%
  triton_mm_2942 0.0258 ms 90.5%
  triton_mm_2950 0.0262 ms 89.2%
  mm 0.0284 ms 82.2%
  triton_mm_2954 0.0300 ms 77.9%
  triton_mm_2941 0.0348 ms 67.0%
  triton_mm_2940 0.0357 ms 65.5%
  triton_mm_2945 0.0378 ms 61.8%
  triton_mm_2949 0.0401 ms 58.2%
  triton_mm_2939 0.0421 ms 55.4%
SingleProcess AUTOTUNE benchmarking takes 2.1657 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_7 0.0241 ms 100.0%
  triton_mm_8 0.0248 ms 97.3%
  triton_mm_4 0.0255 ms 94.7%
  triton_mm_9 0.0258 ms 93.7%
  triton_mm_11 0.0258 ms 93.6%
  triton_mm_12 0.0260 ms 92.9%
  triton_mm_6 0.0260 ms 92.7%
  triton_mm_16 0.0271 ms 88.9%
  triton_mm_2 0.0273 ms 88.4%
  triton_mm_3 0.0273 ms 88.4%
SingleProcess AUTOTUNE benchmarking takes 2.1799 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(12x3840, 12x1920, 1920x3840)
  triton_mm_2872 0.0228 ms 100.0%
  triton_mm_2876 0.0232 ms 98.1%
  triton_mm_2868 0.0236 ms 96.5%
  triton_mm_2880 0.0243 ms 93.9%
  bias_addmm 0.0267 ms 85.5%
  triton_mm_2871 0.0269 ms 84.6%
  triton_mm_2875 0.0279 ms 81.6%
  triton_mm_2866 0.0281 ms 81.0%
  triton_mm_2865 0.0283 ms 80.6%
  triton_mm_2867 0.0289 ms 78.8%
SingleProcess AUTOTUNE benchmarking takes 2.2832 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2741 ms 100.0%
  mm 1.4404 ms 88.5%
  triton_mm_8754 1.5627 ms 81.5%
  triton_mm_8750 1.5740 ms 80.9%
  triton_mm_8758 1.5772 ms 80.8%
  triton_mm_8752 1.6867 ms 75.5%
  triton_mm_8751 1.6886 ms 75.5%
  triton_mm_8755 1.6892 ms 75.4%
  triton_mm_8747 2.0876 ms 61.0%
  triton_mm_8748 2.1504 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.3279 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10431 1.2832 ms 100.0%
  mm 1.3023 ms 98.5%
  triton_mm_10428 1.5584 ms 82.3%
  triton_mm_10432 1.5860 ms 80.9%
  triton_mm_10424 1.5902 ms 80.7%
  triton_mm_10429 1.6993 ms 75.5%
  triton_mm_10425 1.7021 ms 75.4%
  triton_mm_10426 1.7201 ms 74.6%
  triton_mm_10421 2.0995 ms 61.1%
  triton_mm_10422 2.1639 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2245 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3793 ms 100.0%
  mm 0.3878 ms 97.8%
  triton_mm_11504 0.4200 ms 90.3%
  triton_mm_11508 0.4287 ms 88.5%
  triton_mm_11505 0.4512 ms 84.0%
  triton_mm_11506 0.4604 ms 82.4%
  triton_mm_11512 0.4665 ms 81.3%
  triton_mm_11509 0.4724 ms 80.3%
  triton_mm_11501 0.5652 ms 67.1%
  triton_mm_11502 0.5847 ms 64.9%
SingleProcess AUTOTUNE benchmarking takes 2.0683 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3991 ms 100.0%
  mm 0.4134 ms 96.5%
  triton_mm_11523 0.4432 ms 90.1%
  triton_mm_11527 0.4549 ms 87.7%
  triton_mm_11524 0.4756 ms 83.9%
  triton_mm_11525 0.4846 ms 82.4%
  triton_mm_11531 0.4932 ms 80.9%
  triton_mm_11528 0.4977 ms 80.2%
  triton_mm_11520 0.5993 ms 66.6%
  triton_mm_11521 0.6213 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.1446 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0747 ms 100.0%
  triton_mm_8668 0.0758 ms 98.5%
  triton_mm_8659 0.0765 ms 97.6%
  triton_mm_8654 0.0767 ms 97.3%
  triton_mm_8666 0.0772 ms 96.8%
  triton_mm_8653 0.0774 ms 96.4%
  triton_mm_8662 0.0776 ms 96.2%
  triton_mm_8661 0.0782 ms 95.5%
  mm 0.0782 ms 95.5%
  triton_mm_8665 0.0787 ms 94.8%
SingleProcess AUTOTUNE benchmarking takes 1.9862 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0314 ms 100.0%
  triton_mm_8677 0.0314 ms 100.0%
  triton_mm_8678 0.0316 ms 99.5%
  triton_mm_8681 0.0317 ms 98.9%
  triton_mm_8680 0.0328 ms 95.8%
  triton_mm_8676 0.0328 ms 95.7%
  triton_mm_8673 0.0334 ms 93.9%
  triton_mm_8672 0.0349 ms 89.8%
  triton_mm_8684 0.0349 ms 89.8%
  triton_mm_8671 0.0350 ms 89.8%
SingleProcess AUTOTUNE benchmarking takes 2.1979 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9807 ms 100.0%
  triton_mm_8702 0.9842 ms 99.6%
  addmm 1.1091 ms 88.4%
  triton_mm_8699 1.1686 ms 83.9%
  triton_mm_8695 1.1769 ms 83.3%
  triton_mm_8703 1.2289 ms 79.8%
  triton_mm_8700 1.2729 ms 77.0%
  triton_mm_8697 1.2792 ms 76.7%
  triton_mm_8696 1.2812 ms 76.5%
  triton_mm_8692 1.5830 ms 62.0%
SingleProcess AUTOTUNE benchmarking takes 2.3893 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3319 ms 100.0%
  mm 0.3428 ms 96.8%
  triton_mm_8714 0.3982 ms 83.4%
  triton_mm_8722 0.4038 ms 82.2%
  triton_mm_8718 0.4039 ms 82.2%
  triton_mm_8715 0.4330 ms 76.7%
  triton_mm_8716 0.4341 ms 76.5%
  triton_mm_8719 0.4405 ms 75.3%
  triton_mm_8711 0.5312 ms 62.5%
  triton_mm_8712 0.5454 ms 60.9%
SingleProcess AUTOTUNE benchmarking takes 2.1411 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2763 ms 100.0%
  mm 1.2822 ms 99.5%
  triton_mm_8777 1.4908 ms 85.6%
  triton_mm_8769 1.6032 ms 79.6%
  triton_mm_8773 1.6358 ms 78.0%
  triton_mm_8771 1.6648 ms 76.7%
  triton_mm_8770 1.6869 ms 75.7%
  triton_mm_8774 1.7023 ms 75.0%
  triton_mm_8766 2.1285 ms 60.0%
  triton_mm_8767 2.1553 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.3243 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11428 0.0386 ms 100.0%
  triton_mm_11423 0.0392 ms 98.4%
  triton_mm_11424 0.0395 ms 97.7%
  triton_mm_11432 0.0396 ms 97.4%
  triton_mm_11437 0.0398 ms 96.9%
  triton_mm_11422 0.0405 ms 95.3%
  triton_mm_11425 0.0406 ms 95.1%
  triton_mm_11435 0.0408 ms 94.7%
  triton_mm_11431 0.0432 ms 89.3%
  mm 0.0440 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 1.9687 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3895 ms 100.0%
  bias_addmm 5.4024 ms 99.8%
  addmm 6.0663 ms 88.8%
  triton_mm_11489 6.6442 ms 81.1%
  triton_mm_11485 6.6848 ms 80.6%
  triton_mm_11493 6.7003 ms 80.4%
  triton_mm_11490 7.1732 ms 75.1%
  triton_mm_11487 7.2599 ms 74.2%
  triton_mm_11486 7.2773 ms 74.1%
  triton_mm_11482 8.9151 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 3.0488 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1772 ms 100.0%
  bias_addmm 0.1783 ms 99.4%
  triton_mm_11542 0.1802 ms 98.3%
  triton_mm_11546 0.1808 ms 98.0%
  addmm 0.1854 ms 95.6%
  triton_mm_11543 0.1943 ms 91.2%
  triton_mm_11544 0.1956 ms 90.6%
  triton_mm_11547 0.1958 ms 90.5%
  triton_mm_11550 0.2145 ms 82.6%
  triton_mm_11540 0.2370 ms 74.8%
SingleProcess AUTOTUNE benchmarking takes 2.1896 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:56, 139.46s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3942 ms 100.0%
  triton_mm_11523 0.4353 ms 90.6%
  triton_mm_11527 0.4500 ms 87.6%
  mm 0.4563 ms 86.4%
  triton_mm_11524 0.4687 ms 84.1%
  triton_mm_11525 0.4798 ms 82.2%
  triton_mm_11531 0.4888 ms 80.7%
  triton_mm_11528 0.4916 ms 80.2%
  triton_mm_11520 0.5958 ms 66.2%
  triton_mm_11521 0.6121 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.0661 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11428 0.0383 ms 100.0%
  triton_mm_11423 0.0389 ms 98.6%
  triton_mm_11424 0.0394 ms 97.3%
  triton_mm_11437 0.0397 ms 96.5%
  triton_mm_11422 0.0401 ms 95.6%
  triton_mm_11432 0.0402 ms 95.5%
  triton_mm_11435 0.0406 ms 94.3%
  triton_mm_11425 0.0413 ms 92.9%
  triton_mm_11431 0.0439 ms 87.4%
  mm 0.0446 ms 86.0%
SingleProcess AUTOTUNE benchmarking takes 1.9607 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3924 ms 100.0%
  bias_addmm 5.4029 ms 99.8%
  addmm 6.1169 ms 88.2%
  triton_mm_11489 6.5826 ms 81.9%
  triton_mm_11493 6.6998 ms 80.5%
  triton_mm_11485 6.7744 ms 79.6%
  triton_mm_11490 7.1695 ms 75.2%
  triton_mm_11486 7.2792 ms 74.1%
  triton_mm_11487 7.2828 ms 74.0%
  triton_mm_11482 8.8402 ms 61.0%
SingleProcess AUTOTUNE benchmarking takes 3.0780 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1754 ms 100.0%
  bias_addmm 0.1772 ms 99.0%
  triton_mm_11546 0.1789 ms 98.0%
  triton_mm_11542 0.1806 ms 97.1%
  addmm 0.1830 ms 95.8%
  triton_mm_11543 0.1921 ms 91.3%
  triton_mm_11547 0.1934 ms 90.7%
  triton_mm_11544 0.1953 ms 89.8%
  triton_mm_11550 0.2137 ms 82.1%
  triton_mm_11540 0.2344 ms 74.8%
SingleProcess AUTOTUNE benchmarking takes 2.2153 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:57, 139.47s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2555 ms 100.0%
  mm 1.4448 ms 86.9%
  triton_mm_8754 1.5460 ms 81.2%
  triton_mm_8758 1.5567 ms 80.7%
  triton_mm_8750 1.5638 ms 80.3%
  triton_mm_8755 1.6869 ms 74.4%
  triton_mm_8751 1.6885 ms 74.4%
  triton_mm_8752 1.6943 ms 74.1%
  triton_mm_8747 2.0718 ms 60.6%
  triton_mm_8748 2.1532 ms 58.3%
SingleProcess AUTOTUNE benchmarking takes 2.3395 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9129 1.2565 ms 100.0%
  mm 1.2608 ms 99.7%
  triton_mm_9126 1.5461 ms 81.3%
  triton_mm_9130 1.5582 ms 80.6%
  triton_mm_9122 1.5630 ms 80.4%
  triton_mm_9127 1.6681 ms 75.3%
  triton_mm_9123 1.6833 ms 74.6%
  triton_mm_9124 1.6889 ms 74.4%
  triton_mm_9119 2.0708 ms 60.7%
  triton_mm_9120 2.1293 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2244 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3681 ms 100.0%
  mm 0.3812 ms 96.6%
  triton_mm_11504 0.4104 ms 89.7%
  triton_mm_11508 0.4224 ms 87.2%
  triton_mm_11505 0.4456 ms 82.6%
  triton_mm_11506 0.4540 ms 81.1%
  triton_mm_11512 0.4609 ms 79.9%
  triton_mm_11509 0.4650 ms 79.2%
  triton_mm_11501 0.5559 ms 66.2%
  triton_mm_11502 0.5772 ms 63.8%
SingleProcess AUTOTUNE benchmarking takes 2.0617 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3889 ms 100.0%
  mm 0.4019 ms 96.8%
  triton_mm_11523 0.4293 ms 90.6%
  triton_mm_11527 0.4463 ms 87.2%
  triton_mm_11524 0.4657 ms 83.5%
  triton_mm_11525 0.4744 ms 82.0%
  triton_mm_11528 0.4866 ms 79.9%
  triton_mm_11531 0.4904 ms 79.3%
  triton_mm_11520 0.5860 ms 66.4%
  triton_mm_11521 0.6083 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 2.1525 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0750 ms 100.0%
  triton_mm_8668 0.0754 ms 99.5%
  triton_mm_8659 0.0756 ms 99.3%
  triton_mm_8662 0.0762 ms 98.4%
  triton_mm_8653 0.0772 ms 97.1%
  triton_mm_8654 0.0772 ms 97.1%
  mm 0.0777 ms 96.6%
  triton_mm_8661 0.0777 ms 96.6%
  triton_mm_8665 0.0780 ms 96.3%
  triton_mm_8666 0.0783 ms 95.9%
SingleProcess AUTOTUNE benchmarking takes 6.0290 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8680 0.0316 ms 100.0%
  triton_mm_8678 0.0316 ms 99.9%
  triton_mm_8677 0.0317 ms 99.6%
  triton_mm_8675 0.0317 ms 99.5%
  triton_mm_8681 0.0320 ms 98.6%
  triton_mm_8676 0.0329 ms 95.9%
  triton_mm_8673 0.0337 ms 93.8%
  triton_mm_8684 0.0360 ms 87.7%
  triton_mm_8672 0.0361 ms 87.4%
  triton_mm_8671 0.0364 ms 86.7%
SingleProcess AUTOTUNE benchmarking takes 2.2167 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9684 ms 100.0%
  bias_addmm 0.9692 ms 99.9%
  addmm 1.0911 ms 88.8%
  triton_mm_8695 1.1564 ms 83.7%
  triton_mm_8699 1.1689 ms 82.8%
  triton_mm_8703 1.2193 ms 79.4%
  triton_mm_8700 1.2706 ms 76.2%
  triton_mm_8696 1.2728 ms 76.1%
  triton_mm_8697 1.2807 ms 75.6%
  triton_mm_8692 1.5770 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.3958 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3245 ms 100.0%
  mm 0.3348 ms 96.9%
  triton_mm_8714 0.3908 ms 83.0%
  triton_mm_8718 0.3968 ms 81.8%
  triton_mm_8722 0.4024 ms 80.7%
  triton_mm_8715 0.4231 ms 76.7%
  triton_mm_8716 0.4313 ms 75.3%
  triton_mm_8719 0.4324 ms 75.0%
  triton_mm_8711 0.5188 ms 62.6%
  triton_mm_8712 0.5432 ms 59.7%
SingleProcess AUTOTUNE benchmarking takes 2.1462 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2563 ms 100.0%
  mm 1.2664 ms 99.2%
  triton_mm_8777 1.4752 ms 85.2%
  triton_mm_8769 1.5911 ms 79.0%
  triton_mm_8773 1.6251 ms 77.3%
  triton_mm_8771 1.6681 ms 75.3%
  triton_mm_8770 1.6745 ms 75.0%
  triton_mm_8774 1.6892 ms 74.4%
  triton_mm_8766 2.0955 ms 60.0%
  triton_mm_8767 2.1438 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.3407 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8795 0.9653 ms 100.0%
  bias_addmm 0.9700 ms 99.5%
  addmm 1.1030 ms 87.5%
  triton_mm_8792 1.1693 ms 82.6%
  triton_mm_8788 1.1787 ms 81.9%
  triton_mm_8796 1.2172 ms 79.3%
  triton_mm_8793 1.2717 ms 75.9%
  triton_mm_8789 1.2877 ms 75.0%
  triton_mm_8790 1.2899 ms 74.8%
  triton_mm_8785 1.5791 ms 61.1%
SingleProcess AUTOTUNE benchmarking takes 2.3204 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8814 0.3245 ms 100.0%
  mm 0.3346 ms 97.0%
  triton_mm_8807 0.3878 ms 83.7%
  triton_mm_8811 0.3989 ms 81.4%
  triton_mm_8815 0.4012 ms 80.9%
  triton_mm_8808 0.4258 ms 76.2%
  triton_mm_8812 0.4307 ms 75.4%
  triton_mm_8809 0.4310 ms 75.3%
  triton_mm_8804 0.5268 ms 61.6%
  triton_mm_8805 0.5416 ms 59.9%
SingleProcess AUTOTUNE benchmarking takes 2.0791 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8828 0.0750 ms 100.0%
  triton_mm_8833 0.0754 ms 99.4%
  triton_mm_8824 0.0755 ms 99.3%
  triton_mm_8827 0.0764 ms 98.2%
  triton_mm_8818 0.0774 ms 96.9%
  triton_mm_8831 0.0774 ms 96.9%
  triton_mm_8826 0.0777 ms 96.6%
  triton_mm_8819 0.0779 ms 96.3%
  mm 0.0784 ms 95.7%
  triton_mm_8830 0.0792 ms 94.7%
SingleProcess AUTOTUNE benchmarking takes 1.9168 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9695 ms 100.0%
  triton_mm_9911 0.9794 ms 99.0%
  addmm 1.1029 ms 87.9%
  triton_mm_9908 1.1701 ms 82.9%
  triton_mm_9904 1.1762 ms 82.4%
  triton_mm_9912 1.2208 ms 79.4%
  triton_mm_9909 1.2711 ms 76.3%
  triton_mm_9905 1.2882 ms 75.3%
  triton_mm_9906 1.2918 ms 75.0%
  triton_mm_9901 1.5731 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 2.3261 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9930 0.3268 ms 100.0%
  mm 0.3357 ms 97.3%
  triton_mm_9923 0.3910 ms 83.6%
  triton_mm_9927 0.3974 ms 82.2%
  triton_mm_9931 0.4044 ms 80.8%
  triton_mm_9924 0.4280 ms 76.3%
  triton_mm_9928 0.4334 ms 75.4%
  triton_mm_9925 0.4341 ms 75.3%
  triton_mm_9920 0.5240 ms 62.4%
  triton_mm_9921 0.5495 ms 59.5%
SingleProcess AUTOTUNE benchmarking takes 2.0613 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9944 0.0750 ms 100.0%
  triton_mm_9940 0.0756 ms 99.3%
  triton_mm_9949 0.0763 ms 98.4%
  triton_mm_9943 0.0764 ms 98.2%
  triton_mm_9934 0.0776 ms 96.7%
  triton_mm_9947 0.0776 ms 96.6%
  triton_mm_9942 0.0776 ms 96.6%
  mm 0.0777 ms 96.5%
  triton_mm_9935 0.0781 ms 96.0%
  triton_mm_9946 0.0790 ms 94.9%
SingleProcess AUTOTUNE benchmarking takes 1.9051 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  mm 1.2666 ms 100.0%
  triton_mm_9985 1.2671 ms 100.0%
  triton_mm_9986 1.4849 ms 85.3%
  triton_mm_9978 1.5993 ms 79.2%
  triton_mm_9982 1.6331 ms 77.6%
  triton_mm_9980 1.6865 ms 75.1%
  triton_mm_9983 1.6992 ms 74.5%
  triton_mm_9979 1.7197 ms 73.7%
  triton_mm_9975 2.1299 ms 59.5%
  triton_mm_9976 2.1728 ms 58.3%
SingleProcess AUTOTUNE benchmarking takes 2.2373 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11432 0.0391 ms 100.0%
  triton_mm_11423 0.0392 ms 99.8%
  triton_mm_11428 0.0396 ms 98.7%
  triton_mm_11437 0.0397 ms 98.5%
  triton_mm_11424 0.0400 ms 97.8%
  triton_mm_11422 0.0401 ms 97.5%
  triton_mm_11435 0.0408 ms 95.8%
  triton_mm_11425 0.0409 ms 95.5%
  triton_mm_11431 0.0431 ms 90.8%
  mm 0.0446 ms 87.7%
SingleProcess AUTOTUNE benchmarking takes 1.9512 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3903 ms 100.0%
  bias_addmm 5.4222 ms 99.4%
  addmm 6.0622 ms 88.9%
  triton_mm_11489 6.5972 ms 81.7%
  triton_mm_11485 6.7150 ms 80.3%
  triton_mm_11493 6.7226 ms 80.2%
  triton_mm_11490 7.1965 ms 74.9%
  triton_mm_11487 7.2132 ms 74.7%
  triton_mm_11486 7.2145 ms 74.7%
  triton_mm_11482 8.8850 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 3.0681 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1750 ms 100.0%
  bias_addmm 0.1769 ms 98.9%
  triton_mm_11546 0.1773 ms 98.7%
  triton_mm_11542 0.1784 ms 98.1%
  addmm 0.1834 ms 95.4%
  triton_mm_11543 0.1909 ms 91.7%
  triton_mm_11547 0.1928 ms 90.8%
  triton_mm_11544 0.1957 ms 89.4%
  triton_mm_11550 0.2142 ms 81.7%
  triton_mm_11540 0.2368 ms 73.9%
SingleProcess AUTOTUNE benchmarking takes 2.1860 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:57, 139.48s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2468 ms 100.0%
  mm 1.4436 ms 86.4%
  triton_mm_8754 1.5268 ms 81.7%
  triton_mm_8758 1.5382 ms 81.1%
  triton_mm_8750 1.5411 ms 80.9%
  triton_mm_8755 1.6480 ms 75.7%
  triton_mm_8751 1.6686 ms 74.7%
  triton_mm_8752 1.6696 ms 74.7%
  triton_mm_8747 2.0462 ms 60.9%
  triton_mm_8748 2.1018 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.3304 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9408 1.2459 ms 100.0%
  mm 1.2781 ms 97.5%
  triton_mm_9405 1.5232 ms 81.8%
  triton_mm_9409 1.5341 ms 81.2%
  triton_mm_9401 1.5404 ms 80.9%
  triton_mm_9406 1.6622 ms 75.0%
  triton_mm_9402 1.6640 ms 74.9%
  triton_mm_9403 1.6848 ms 73.9%
  triton_mm_9398 2.0661 ms 60.3%
  triton_mm_9399 2.1202 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.2341 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10338 1.2539 ms 100.0%
  mm 1.2787 ms 98.1%
  triton_mm_10335 1.5241 ms 82.3%
  triton_mm_10331 1.5414 ms 81.3%
  triton_mm_10339 1.5517 ms 80.8%
  triton_mm_10336 1.6619 ms 75.4%
  triton_mm_10332 1.6642 ms 75.3%
  triton_mm_10333 1.6846 ms 74.4%
  triton_mm_10328 2.0628 ms 60.8%
  triton_mm_10329 2.1199 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2447 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10617 1.2557 ms 100.0%
  mm 1.2804 ms 98.1%
  triton_mm_10614 1.5389 ms 81.6%
  triton_mm_10610 1.5412 ms 81.5%
  triton_mm_10618 1.5524 ms 80.9%
  triton_mm_10615 1.6637 ms 75.5%
  triton_mm_10611 1.6651 ms 75.4%
  triton_mm_10612 1.6834 ms 74.6%
  triton_mm_10607 2.0632 ms 60.9%
  triton_mm_10608 2.1194 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2364 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10803 1.2561 ms 100.0%
  mm 1.2816 ms 98.0%
  triton_mm_10800 1.5287 ms 82.2%
  triton_mm_10804 1.5519 ms 80.9%
  triton_mm_10796 1.5540 ms 80.8%
  triton_mm_10801 1.6629 ms 75.5%
  triton_mm_10797 1.6652 ms 75.4%
  triton_mm_10798 1.6835 ms 74.6%
  triton_mm_10793 2.0634 ms 60.9%
  triton_mm_10794 2.1188 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2276 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10989 1.2576 ms 100.0%
  mm 1.2860 ms 97.8%
  triton_mm_10986 1.5393 ms 81.7%
  triton_mm_10982 1.5426 ms 81.5%
  triton_mm_10990 1.5592 ms 80.7%
  triton_mm_10987 1.6633 ms 75.6%
  triton_mm_10983 1.6654 ms 75.5%
  triton_mm_10984 1.6820 ms 74.8%
  triton_mm_10979 2.0635 ms 60.9%
  triton_mm_10980 2.1207 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2293 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11454 1.2564 ms 100.0%
  mm 1.2809 ms 98.1%
  triton_mm_11451 1.5257 ms 82.3%
  triton_mm_11455 1.5527 ms 80.9%
  triton_mm_11447 1.5540 ms 80.8%
  triton_mm_11452 1.6631 ms 75.5%
  triton_mm_11448 1.6664 ms 75.4%
  triton_mm_11449 1.6817 ms 74.7%
  triton_mm_11444 2.0634 ms 60.9%
  triton_mm_11445 2.1199 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2255 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3711 ms 100.0%
  mm 0.3812 ms 97.3%
  triton_mm_11504 0.4131 ms 89.8%
  triton_mm_11508 0.4188 ms 88.6%
  triton_mm_11505 0.4410 ms 84.2%
  triton_mm_11506 0.4546 ms 81.6%
  triton_mm_11512 0.4591 ms 80.8%
  triton_mm_11509 0.4641 ms 80.0%
  triton_mm_11501 0.5585 ms 66.4%
  triton_mm_11502 0.5730 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 2.1782 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3923 ms 100.0%
  mm 0.4041 ms 97.1%
  triton_mm_11523 0.4359 ms 90.0%
  triton_mm_11527 0.4477 ms 87.6%
  triton_mm_11524 0.4671 ms 84.0%
  triton_mm_11525 0.4775 ms 82.1%
  triton_mm_11531 0.4869 ms 80.6%
  triton_mm_11528 0.4919 ms 79.7%
  triton_mm_11520 0.5896 ms 66.5%
  triton_mm_11521 0.6077 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.0863 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0318 ms 100.0%
  triton_mm_8676 0.0319 ms 99.9%
  triton_mm_8681 0.0321 ms 99.2%
  triton_mm_8678 0.0329 ms 96.8%
  triton_mm_8680 0.0329 ms 96.8%
  triton_mm_8677 0.0330 ms 96.5%
  triton_mm_8673 0.0347 ms 91.8%
  triton_mm_8684 0.0353 ms 90.1%
  triton_mm_8671 0.0362 ms 88.0%
  triton_mm_8672 0.0362 ms 88.0%
SingleProcess AUTOTUNE benchmarking takes 2.2318 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:57, 139.48s/it]AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3690 ms 100.0%
  triton_mm_11504 0.4139 ms 89.2%
  triton_mm_11508 0.4178 ms 88.3%
  mm 0.4335 ms 85.1%
  triton_mm_11505 0.4409 ms 83.7%
  triton_mm_11506 0.4587 ms 80.4%
  triton_mm_11512 0.4588 ms 80.4%
  triton_mm_11509 0.4591 ms 80.4%
  triton_mm_11501 0.5571 ms 66.2%
  triton_mm_11502 0.5760 ms 64.1%
SingleProcess AUTOTUNE benchmarking takes 2.1417 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3891 ms 100.0%
  mm 0.4021 ms 96.8%
  triton_mm_11523 0.4325 ms 90.0%
  triton_mm_11527 0.4452 ms 87.4%
  triton_mm_11524 0.4660 ms 83.5%
  triton_mm_11525 0.4735 ms 82.2%
  triton_mm_11531 0.4879 ms 79.8%
  triton_mm_11528 0.4889 ms 79.6%
  triton_mm_11520 0.5880 ms 66.2%
  triton_mm_11521 0.6048 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.0680 seconds and 0.0009 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0741 ms 100.0%
  triton_mm_8668 0.0753 ms 98.4%
  triton_mm_8659 0.0756 ms 98.1%
  triton_mm_8662 0.0760 ms 97.5%
  triton_mm_8653 0.0770 ms 96.3%
  triton_mm_8666 0.0774 ms 95.7%
  triton_mm_8654 0.0779 ms 95.2%
  mm 0.0783 ms 94.7%
  triton_mm_8661 0.0783 ms 94.6%
  triton_mm_8665 0.0788 ms 94.1%
SingleProcess AUTOTUNE benchmarking takes 1.9756 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8677 0.0317 ms 100.0%
  triton_mm_8681 0.0319 ms 99.2%
  triton_mm_8678 0.0329 ms 96.4%
  triton_mm_8675 0.0329 ms 96.2%
  triton_mm_8680 0.0329 ms 96.2%
  triton_mm_8676 0.0332 ms 95.6%
  triton_mm_8673 0.0348 ms 91.0%
  triton_mm_8671 0.0351 ms 90.3%
  triton_mm_8672 0.0363 ms 87.3%
  triton_mm_8684 0.0363 ms 87.3%
SingleProcess AUTOTUNE benchmarking takes 2.1965 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9683 ms 100.0%
  bias_addmm 0.9723 ms 99.6%
  addmm 1.0924 ms 88.6%
  triton_mm_8699 1.1447 ms 84.6%
  triton_mm_8695 1.1643 ms 83.2%
  triton_mm_8703 1.2021 ms 80.6%
  triton_mm_8700 1.2534 ms 77.3%
  triton_mm_8696 1.2550 ms 77.2%
  triton_mm_8697 1.2688 ms 76.3%
  triton_mm_8692 1.5579 ms 62.2%
SingleProcess AUTOTUNE benchmarking takes 2.4220 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3252 ms 100.0%
  mm 0.3360 ms 96.8%
  triton_mm_8714 0.3894 ms 83.5%
  triton_mm_8718 0.3948 ms 82.4%
  triton_mm_8722 0.4002 ms 81.3%
  triton_mm_8715 0.4269 ms 76.2%
  triton_mm_8716 0.4295 ms 75.7%
  triton_mm_8719 0.4296 ms 75.7%
  triton_mm_8711 0.5161 ms 63.0%
  triton_mm_8712 0.5366 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.1515 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2489 ms 100.0%
  mm 1.2647 ms 98.7%
  triton_mm_8777 1.4695 ms 85.0%
  triton_mm_8769 1.5873 ms 78.7%
  triton_mm_8773 1.6110 ms 77.5%
  triton_mm_8770 1.6608 ms 75.2%
  triton_mm_8771 1.6667 ms 74.9%
  triton_mm_8774 1.6757 ms 74.5%
  triton_mm_8766 2.1014 ms 59.4%
  triton_mm_8767 2.1395 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 2.3214 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9647 ms 100.0%
  triton_mm_9167 0.9683 ms 99.6%
  addmm 1.0879 ms 88.7%
  triton_mm_9164 1.1632 ms 82.9%
  triton_mm_9160 1.1649 ms 82.8%
  triton_mm_9168 1.2046 ms 80.1%
  triton_mm_9165 1.2672 ms 76.1%
  triton_mm_9162 1.2829 ms 75.2%
  triton_mm_9161 1.2836 ms 75.2%
  triton_mm_9157 1.5582 ms 61.9%
SingleProcess AUTOTUNE benchmarking takes 2.3093 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9186 0.3261 ms 100.0%
  mm 0.3359 ms 97.1%
  triton_mm_9179 0.3898 ms 83.7%
  triton_mm_9183 0.3980 ms 81.9%
  triton_mm_9187 0.4034 ms 80.8%
  triton_mm_9180 0.4280 ms 76.2%
  triton_mm_9181 0.4312 ms 75.6%
  triton_mm_9184 0.4334 ms 75.2%
  triton_mm_9176 0.5205 ms 62.6%
  triton_mm_9177 0.5413 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 2.0556 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9200 0.0749 ms 100.0%
  triton_mm_9205 0.0753 ms 99.5%
  triton_mm_9199 0.0761 ms 98.5%
  triton_mm_9196 0.0764 ms 98.1%
  triton_mm_9203 0.0774 ms 96.9%
  triton_mm_9190 0.0779 ms 96.2%
  triton_mm_9191 0.0779 ms 96.2%
  triton_mm_9202 0.0781 ms 96.0%
  triton_mm_9198 0.0781 ms 95.9%
  mm 0.0784 ms 95.6%
SingleProcess AUTOTUNE benchmarking takes 1.8965 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9241 1.2580 ms 100.0%
  mm 1.2673 ms 99.3%
  triton_mm_9242 1.4707 ms 85.5%
  triton_mm_9234 1.5998 ms 78.6%
  triton_mm_9238 1.6228 ms 77.5%
  triton_mm_9239 1.6772 ms 75.0%
  triton_mm_9236 1.6849 ms 74.7%
  triton_mm_9235 1.6941 ms 74.3%
  triton_mm_9231 2.1009 ms 59.9%
  triton_mm_9232 2.1400 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.2396 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  mm 1.2572 ms 100.0%
  triton_mm_9613 1.2585 ms 99.9%
  triton_mm_9614 1.4722 ms 85.4%
  triton_mm_9606 1.6007 ms 78.5%
  triton_mm_9610 1.6198 ms 77.6%
  triton_mm_9608 1.6849 ms 74.6%
  triton_mm_9611 1.6932 ms 74.2%
  triton_mm_9607 1.6944 ms 74.2%
  triton_mm_9603 2.1185 ms 59.3%
  triton_mm_9604 2.1399 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2381 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9667 ms 100.0%
  triton_mm_9632 0.9708 ms 99.6%
  addmm 1.1004 ms 87.9%
  triton_mm_9629 1.1697 ms 82.6%
  triton_mm_9625 1.1764 ms 82.2%
  triton_mm_9633 1.2264 ms 78.8%
  triton_mm_9630 1.2708 ms 76.1%
  triton_mm_9626 1.2845 ms 75.3%
  triton_mm_9627 1.2908 ms 74.9%
  triton_mm_9622 1.5730 ms 61.5%
SingleProcess AUTOTUNE benchmarking takes 2.3163 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9651 0.3270 ms 100.0%
  mm 0.3366 ms 97.1%
  triton_mm_9644 0.3904 ms 83.8%
  triton_mm_9648 0.3978 ms 82.2%
  triton_mm_9652 0.4018 ms 81.4%
  triton_mm_9645 0.4276 ms 76.5%
  triton_mm_9646 0.4306 ms 75.9%
  triton_mm_9649 0.4327 ms 75.6%
  triton_mm_9641 0.5209 ms 62.8%
  triton_mm_9642 0.5423 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 2.0582 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9665 0.0742 ms 100.0%
  triton_mm_9670 0.0753 ms 98.5%
  triton_mm_9661 0.0764 ms 97.1%
  triton_mm_9664 0.0770 ms 96.4%
  triton_mm_9668 0.0775 ms 95.8%
  triton_mm_9655 0.0779 ms 95.3%
  triton_mm_9656 0.0779 ms 95.3%
  triton_mm_9667 0.0781 ms 95.0%
  triton_mm_9663 0.0782 ms 94.9%
  mm 0.0783 ms 94.8%
SingleProcess AUTOTUNE benchmarking takes 1.8961 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_11153 0.0741 ms 100.0%
  triton_mm_11149 0.0756 ms 98.0%
  triton_mm_11158 0.0760 ms 97.6%
  triton_mm_11152 0.0761 ms 97.4%
  triton_mm_11143 0.0770 ms 96.3%
  triton_mm_11144 0.0772 ms 96.1%
  triton_mm_11151 0.0773 ms 95.9%
  mm 0.0777 ms 95.4%
  triton_mm_11155 0.0781 ms 94.9%
  triton_mm_11156 0.0784 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.8963 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11424 0.0389 ms 100.0%
  triton_mm_11432 0.0394 ms 98.8%
  triton_mm_11423 0.0395 ms 98.5%
  triton_mm_11437 0.0397 ms 98.0%
  triton_mm_11422 0.0398 ms 97.7%
  triton_mm_11428 0.0401 ms 97.1%
  triton_mm_11425 0.0406 ms 95.8%
  triton_mm_11435 0.0423 ms 92.1%
  triton_mm_11431 0.0433 ms 89.9%
  mm 0.0443 ms 87.8%
SingleProcess AUTOTUNE benchmarking takes 1.9440 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3276 ms 100.0%
  bias_addmm 5.4802 ms 97.2%
  addmm 6.1132 ms 87.1%
  triton_mm_11489 6.5062 ms 81.9%
  triton_mm_11493 6.6252 ms 80.4%
  triton_mm_11485 6.7000 ms 79.5%
  triton_mm_11490 7.1646 ms 74.4%
  triton_mm_11486 7.1939 ms 74.1%
  triton_mm_11487 7.2107 ms 73.9%
  triton_mm_11482 8.8570 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 3.0449 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1744 ms 100.0%
  bias_addmm 0.1768 ms 98.7%
  triton_mm_11546 0.1782 ms 97.9%
  triton_mm_11542 0.1787 ms 97.6%
  addmm 0.1818 ms 95.9%
  triton_mm_11543 0.1907 ms 91.5%
  triton_mm_11547 0.1930 ms 90.4%
  triton_mm_11544 0.1943 ms 89.8%
  triton_mm_11550 0.2133 ms 81.8%
  triton_mm_11540 0.2344 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 2.1968 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:57, 139.48s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2681 ms 100.0%
  mm 1.4508 ms 87.4%
  triton_mm_8754 1.5566 ms 81.5%
  triton_mm_8758 1.5652 ms 81.0%
  triton_mm_8750 1.5737 ms 80.6%
  triton_mm_8755 1.6811 ms 75.4%
  triton_mm_8751 1.6811 ms 75.4%
  triton_mm_8752 1.7046 ms 74.4%
  triton_mm_8747 2.0859 ms 60.8%
  triton_mm_8748 2.1435 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.3311 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9780 1.2825 ms 100.0%
  mm 1.3003 ms 98.6%
  triton_mm_9777 1.5666 ms 81.9%
  triton_mm_9781 1.5844 ms 81.0%
  triton_mm_9773 1.5936 ms 80.5%
  triton_mm_9778 1.6910 ms 75.8%
  triton_mm_9774 1.7000 ms 75.4%
  triton_mm_9775 1.7265 ms 74.3%
  triton_mm_9770 2.0863 ms 61.5%
  triton_mm_9771 2.1909 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2360 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10245 1.2827 ms 100.0%
  mm 1.3023 ms 98.5%
  triton_mm_10242 1.5563 ms 82.4%
  triton_mm_10246 1.5728 ms 81.6%
  triton_mm_10238 1.6107 ms 79.6%
  triton_mm_10243 1.7008 ms 75.4%
  triton_mm_10239 1.7197 ms 74.6%
  triton_mm_10240 1.7252 ms 74.4%
  triton_mm_10235 2.1062 ms 60.9%
  triton_mm_10236 2.1919 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.2279 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3752 ms 100.0%
  mm 0.3836 ms 97.8%
  triton_mm_11504 0.4173 ms 89.9%
  triton_mm_11508 0.4273 ms 87.8%
  triton_mm_11505 0.4485 ms 83.6%
  triton_mm_11506 0.4564 ms 82.2%
  triton_mm_11512 0.4644 ms 80.8%
  triton_mm_11509 0.4693 ms 79.9%
  triton_mm_11501 0.5620 ms 66.8%
  triton_mm_11502 0.5793 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 2.1670 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3975 ms 100.0%
  mm 0.4098 ms 97.0%
  triton_mm_11523 0.4400 ms 90.3%
  triton_mm_11527 0.4542 ms 87.5%
  triton_mm_11524 0.4760 ms 83.5%
  triton_mm_11525 0.4841 ms 82.1%
  triton_mm_11531 0.4923 ms 80.7%
  triton_mm_11528 0.4981 ms 79.8%
  triton_mm_11520 0.6025 ms 66.0%
  triton_mm_11521 0.6190 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.0747 seconds and 0.0008 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0741 ms 100.0%
  triton_mm_8659 0.0758 ms 97.8%
  triton_mm_8668 0.0759 ms 97.7%
  triton_mm_8662 0.0772 ms 96.1%
  triton_mm_8653 0.0774 ms 95.8%
  triton_mm_8661 0.0778 ms 95.3%
  triton_mm_8666 0.0780 ms 95.0%
  triton_mm_8654 0.0781 ms 94.9%
  mm 0.0784 ms 94.6%
  triton_mm_8665 0.0784 ms 94.5%
SingleProcess AUTOTUNE benchmarking takes 1.9836 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0315 ms 100.0%
  triton_mm_8677 0.0316 ms 99.8%
  triton_mm_8678 0.0317 ms 99.4%
  triton_mm_8680 0.0317 ms 99.3%
  triton_mm_8681 0.0319 ms 98.9%
  triton_mm_8676 0.0327 ms 96.3%
  triton_mm_8673 0.0345 ms 91.4%
  triton_mm_8671 0.0350 ms 90.0%
  triton_mm_8672 0.0360 ms 87.6%
  triton_mm_8684 0.0360 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.2200 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9753 ms 100.0%
  triton_mm_8702 0.9785 ms 99.7%
  addmm 1.1048 ms 88.3%
  triton_mm_8695 1.1752 ms 83.0%
  triton_mm_8699 1.1769 ms 82.9%
  triton_mm_8703 1.2307 ms 79.3%
  triton_mm_8700 1.2810 ms 76.1%
  triton_mm_8696 1.2810 ms 76.1%
  triton_mm_8697 1.2820 ms 76.1%
  triton_mm_8692 1.5892 ms 61.4%
SingleProcess AUTOTUNE benchmarking takes 2.4053 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3299 ms 100.0%
  mm 0.3392 ms 97.3%
  triton_mm_8714 0.3948 ms 83.6%
  triton_mm_8718 0.4010 ms 82.3%
  triton_mm_8722 0.4038 ms 81.7%
  triton_mm_8715 0.4319 ms 76.4%
  triton_mm_8716 0.4320 ms 76.4%
  triton_mm_8719 0.4364 ms 75.6%
  triton_mm_8711 0.5250 ms 62.8%
  triton_mm_8712 0.5459 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.1499 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2692 ms 100.0%
  mm 1.2786 ms 99.3%
  triton_mm_8777 1.4852 ms 85.5%
  triton_mm_8769 1.6136 ms 78.7%
  triton_mm_8773 1.6343 ms 77.7%
  triton_mm_8771 1.6755 ms 75.8%
  triton_mm_8774 1.6919 ms 75.0%
  triton_mm_8770 1.6952 ms 74.9%
  triton_mm_8766 2.1179 ms 59.9%
  triton_mm_8767 2.1557 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.3284 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0383 ms 100.0%
  triton_mm_11424 0.0396 ms 96.6%
  triton_mm_11428 0.0396 ms 96.5%
  triton_mm_11422 0.0397 ms 96.5%
  triton_mm_11432 0.0397 ms 96.5%
  triton_mm_11437 0.0405 ms 94.5%
  triton_mm_11435 0.0410 ms 93.4%
  triton_mm_11425 0.0415 ms 92.3%
  triton_mm_11431 0.0444 ms 86.2%
  mm 0.0448 ms 85.4%
SingleProcess AUTOTUNE benchmarking takes 1.9501 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  bias_addmm 5.2956 ms 100.0%
  triton_mm_11492 5.3922 ms 98.2%
  addmm 6.1191 ms 86.5%
  triton_mm_11489 6.6548 ms 79.6%
  triton_mm_11485 6.6813 ms 79.3%
  triton_mm_11493 6.7064 ms 79.0%
  triton_mm_11490 7.1732 ms 73.8%
  triton_mm_11486 7.2730 ms 72.8%
  triton_mm_11487 7.2799 ms 72.7%
  triton_mm_11482 8.8225 ms 60.0%
SingleProcess AUTOTUNE benchmarking takes 6.3507 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1762 ms 100.0%
  bias_addmm 0.1769 ms 99.6%
  triton_mm_11546 0.1803 ms 97.7%
  triton_mm_11542 0.1808 ms 97.4%
  addmm 0.1836 ms 95.9%
  triton_mm_11543 0.1922 ms 91.6%
  triton_mm_11547 0.1935 ms 91.0%
  triton_mm_11544 0.1969 ms 89.5%
  triton_mm_11550 0.2145 ms 82.1%
  triton_mm_11540 0.2347 ms 75.1%
SingleProcess AUTOTUNE benchmarking takes 2.1965 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:57, 139.49s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2668 ms 100.0%
  mm 1.4425 ms 87.8%
  triton_mm_8754 1.5381 ms 82.4%
  triton_mm_8758 1.5661 ms 80.9%
  triton_mm_8750 1.5728 ms 80.5%
  triton_mm_8755 1.6602 ms 76.3%
  triton_mm_8751 1.6822 ms 75.3%
  triton_mm_8752 1.7015 ms 74.5%
  triton_mm_8747 2.0656 ms 61.3%
  triton_mm_8748 2.1419 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.3348 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9501 1.2800 ms 100.0%
  mm 1.2914 ms 99.1%
  triton_mm_9498 1.5608 ms 82.0%
  triton_mm_9502 1.5685 ms 81.6%
  triton_mm_9494 1.5925 ms 80.4%
  triton_mm_9499 1.6810 ms 76.1%
  triton_mm_9495 1.6843 ms 76.0%
  triton_mm_9496 1.7100 ms 74.9%
  triton_mm_9491 2.0932 ms 61.2%
  triton_mm_9492 2.1720 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.2381 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3718 ms 100.0%
  mm 0.3814 ms 97.5%
  triton_mm_11504 0.4110 ms 90.5%
  triton_mm_11508 0.4183 ms 88.9%
  triton_mm_11505 0.4466 ms 83.3%
  triton_mm_11506 0.4535 ms 82.0%
  triton_mm_11512 0.4620 ms 80.5%
  triton_mm_11509 0.4662 ms 79.7%
  triton_mm_11501 0.5591 ms 66.5%
  triton_mm_11502 0.5740 ms 64.8%
SingleProcess AUTOTUNE benchmarking takes 2.1758 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3952 ms 100.0%
  mm 0.4044 ms 97.7%
  triton_mm_11523 0.4382 ms 90.2%
  triton_mm_11527 0.4459 ms 88.6%
  triton_mm_11524 0.4697 ms 84.1%
  triton_mm_11525 0.4821 ms 82.0%
  triton_mm_11531 0.4890 ms 80.8%
  triton_mm_11528 0.4945 ms 79.9%
  triton_mm_11520 0.5947 ms 66.5%
  triton_mm_11521 0.6172 ms 64.0%
SingleProcess AUTOTUNE benchmarking takes 2.0722 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0750 ms 100.0%
  triton_mm_8668 0.0755 ms 99.4%
  triton_mm_8659 0.0763 ms 98.4%
  triton_mm_8654 0.0770 ms 97.5%
  triton_mm_8662 0.0771 ms 97.3%
  triton_mm_8653 0.0780 ms 96.2%
  triton_mm_8666 0.0783 ms 95.9%
  triton_mm_8661 0.0785 ms 95.6%
  mm 0.0787 ms 95.4%
  triton_mm_8665 0.0787 ms 95.3%
SingleProcess AUTOTUNE benchmarking takes 1.9964 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0318 ms 100.0%
  triton_mm_8677 0.0318 ms 100.0%
  triton_mm_8676 0.0329 ms 96.8%
  triton_mm_8678 0.0329 ms 96.7%
  triton_mm_8680 0.0329 ms 96.7%
  triton_mm_8681 0.0333 ms 95.5%
  triton_mm_8673 0.0347 ms 91.8%
  triton_mm_8684 0.0351 ms 90.6%
  triton_mm_8672 0.0362 ms 87.9%
  triton_mm_8671 0.0364 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.2341 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9684 ms 100.0%
  triton_mm_8702 0.9779 ms 99.0%
  addmm 1.1016 ms 87.9%
  triton_mm_8699 1.1644 ms 83.2%
  triton_mm_8695 1.1735 ms 82.5%
  triton_mm_8703 1.2312 ms 78.7%
  triton_mm_8700 1.2657 ms 76.5%
  triton_mm_8696 1.2688 ms 76.3%
  triton_mm_8697 1.2848 ms 75.4%
  triton_mm_8692 1.5720 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 2.3986 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3273 ms 100.0%
  mm 0.3378 ms 96.9%
  triton_mm_8714 0.3907 ms 83.8%
  triton_mm_8718 0.3962 ms 82.6%
  triton_mm_8722 0.4022 ms 81.4%
  triton_mm_8715 0.4257 ms 76.9%
  triton_mm_8719 0.4345 ms 75.3%
  triton_mm_8716 0.4349 ms 75.3%
  triton_mm_8711 0.5204 ms 62.9%
  triton_mm_8712 0.5418 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 2.1845 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2548 ms 100.0%
  mm 1.2646 ms 99.2%
  triton_mm_8777 1.4884 ms 84.3%
  triton_mm_8769 1.5973 ms 78.6%
  triton_mm_8773 1.6138 ms 77.8%
  triton_mm_8771 1.6589 ms 75.6%
  triton_mm_8770 1.6753 ms 74.9%
  triton_mm_8774 1.7101 ms 73.4%
  triton_mm_8766 2.1140 ms 59.4%
  triton_mm_8767 2.1580 ms 58.1%
SingleProcess AUTOTUNE benchmarking takes 2.3209 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0387 ms 100.0%
  triton_mm_11424 0.0393 ms 98.4%
  triton_mm_11432 0.0398 ms 97.1%
  triton_mm_11437 0.0401 ms 96.4%
  triton_mm_11428 0.0402 ms 96.3%
  triton_mm_11422 0.0403 ms 96.0%
  triton_mm_11425 0.0418 ms 92.6%
  triton_mm_11435 0.0422 ms 91.7%
  triton_mm_11431 0.0444 ms 87.0%
  mm 0.0451 ms 85.8%
SingleProcess AUTOTUNE benchmarking takes 1.9664 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  bias_addmm 5.2320 ms 100.0%
  triton_mm_11492 5.3973 ms 96.9%
  addmm 6.0514 ms 86.5%
  triton_mm_11489 6.5726 ms 79.6%
  triton_mm_11493 6.7113 ms 78.0%
  triton_mm_11485 6.7618 ms 77.4%
  triton_mm_11490 7.1636 ms 73.0%
  triton_mm_11486 7.2026 ms 72.6%
  triton_mm_11487 7.2790 ms 71.9%
  triton_mm_11482 8.9604 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 3.0828 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1764 ms 100.0%
  bias_addmm 0.1780 ms 99.1%
  triton_mm_11546 0.1794 ms 98.3%
  triton_mm_11542 0.1796 ms 98.2%
  addmm 0.1832 ms 96.3%
  triton_mm_11543 0.1917 ms 92.0%
  triton_mm_11547 0.1923 ms 91.7%
  triton_mm_11544 0.1965 ms 89.7%
  triton_mm_11550 0.2150 ms 82.0%
  triton_mm_11540 0.2356 ms 74.9%
SingleProcess AUTOTUNE benchmarking takes 2.2076 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:39:57, 139.49s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3962 ms 100.0%
  triton_mm_11523 0.4364 ms 90.8%
  triton_mm_11527 0.4523 ms 87.6%
  mm 0.4573 ms 86.6%
  triton_mm_11524 0.4710 ms 84.1%
  triton_mm_11525 0.4789 ms 82.7%
  triton_mm_11531 0.4883 ms 81.1%
  triton_mm_11528 0.4929 ms 80.4%
  triton_mm_11520 0.5915 ms 67.0%
  triton_mm_11521 0.6144 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.0852 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9676 ms 100.0%
  triton_mm_8702 0.9736 ms 99.4%
  addmm 1.1025 ms 87.8%
  triton_mm_8699 1.1631 ms 83.2%
  triton_mm_8695 1.1681 ms 82.8%
  triton_mm_8703 1.2191 ms 79.4%
  triton_mm_8700 1.2679 ms 76.3%
  triton_mm_8696 1.2692 ms 76.2%
  triton_mm_8697 1.2783 ms 75.7%
  triton_mm_8692 1.5725 ms 61.5%
SingleProcess AUTOTUNE benchmarking takes 2.4249 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3286 ms 100.0%
  mm 0.3383 ms 97.1%
  triton_mm_8714 0.3954 ms 83.1%
  triton_mm_8718 0.4013 ms 81.9%
  triton_mm_8722 0.4037 ms 81.4%
  triton_mm_8715 0.4301 ms 76.4%
  triton_mm_8716 0.4330 ms 75.9%
  triton_mm_8719 0.4352 ms 75.5%
  triton_mm_8711 0.5244 ms 62.7%
  triton_mm_8712 0.5435 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.1757 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2633 ms 100.0%
  mm 1.2708 ms 99.4%
  triton_mm_8777 1.4764 ms 85.6%
  triton_mm_8769 1.6052 ms 78.7%
  triton_mm_8773 1.6250 ms 77.7%
  triton_mm_8771 1.6627 ms 76.0%
  triton_mm_8770 1.6812 ms 75.1%
  triton_mm_8774 1.6928 ms 74.6%
  triton_mm_8766 2.1212 ms 59.6%
  triton_mm_8767 2.1340 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.3462 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_11418 0.3292 ms 100.0%
  mm 0.3368 ms 97.8%
  triton_mm_11411 0.3952 ms 83.3%
  triton_mm_11415 0.4006 ms 82.2%
  triton_mm_11419 0.4047 ms 81.3%
  triton_mm_11412 0.4315 ms 76.3%
  triton_mm_11413 0.4344 ms 75.8%
  triton_mm_11416 0.4346 ms 75.7%
  triton_mm_11408 0.5241 ms 62.8%
  triton_mm_11409 0.5436 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.0751 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11428 0.0387 ms 100.0%
  triton_mm_11424 0.0388 ms 99.7%
  triton_mm_11423 0.0394 ms 98.1%
  triton_mm_11437 0.0398 ms 97.2%
  triton_mm_11422 0.0400 ms 96.7%
  triton_mm_11432 0.0408 ms 94.8%
  triton_mm_11425 0.0409 ms 94.5%
  triton_mm_11435 0.0419 ms 92.2%
  triton_mm_11431 0.0442 ms 87.5%
  mm 0.0446 ms 86.7%
SingleProcess AUTOTUNE benchmarking takes 1.9807 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  mm 1.2672 ms 100.0%
  triton_mm_11473 1.2690 ms 99.9%
  triton_mm_11474 1.4841 ms 85.4%
  triton_mm_11466 1.6021 ms 79.1%
  triton_mm_11470 1.6312 ms 77.7%
  triton_mm_11468 1.6751 ms 75.6%
  triton_mm_11467 1.6817 ms 75.4%
  triton_mm_11471 1.6920 ms 74.9%
  triton_mm_11463 2.1340 ms 59.4%
  triton_mm_11464 2.1362 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.2538 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3869 ms 100.0%
  bias_addmm 5.4044 ms 99.7%
  addmm 6.0553 ms 89.0%
  triton_mm_11489 6.5735 ms 81.9%
  triton_mm_11485 6.6933 ms 80.5%
  triton_mm_11493 6.6937 ms 80.5%
  triton_mm_11490 7.1662 ms 75.2%
  triton_mm_11486 7.1921 ms 74.9%
  triton_mm_11487 7.1956 ms 74.9%
  triton_mm_11482 8.9301 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 3.0685 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1753 ms 100.0%
  bias_addmm 0.1769 ms 99.1%
  triton_mm_11546 0.1794 ms 97.7%
  triton_mm_11542 0.1805 ms 97.1%
  addmm 0.1834 ms 95.6%
  triton_mm_11543 0.1936 ms 90.5%
  triton_mm_11547 0.1947 ms 90.0%
  triton_mm_11544 0.1969 ms 89.0%
  triton_mm_11550 0.2137 ms 82.0%
  triton_mm_11540 0.2365 ms 74.1%
SingleProcess AUTOTUNE benchmarking takes 2.2296 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:19<1:40:02, 139.59s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.62s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.62s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.63s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.63s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.63s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.63s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:44, 59.63s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:23<41:46, 59.67s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:33, 33.01s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:24<22:34, 33.04s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.43s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.44s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.44s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.44s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.44s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.44s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:37, 20.44s/it]
Validation Progress:   9%|9         | 4/44 [02:25<13:38, 20.45s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.46s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.47s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.47s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.47s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.47s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.47s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.47s/it]
Validation Progress:  11%|#1        | 5/44 [02:26<08:45, 13.48s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:27<05:52,  9.28s/it]
Validation Progress:  14%|#3        | 6/44 [02:28<05:52,  9.29s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.63s/it]
Validation Progress:  16%|#5        | 7/44 [02:29<04:05,  6.64s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  18%|#8        | 8/44 [02:30<02:56,  4.90s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.75s/it]
Validation Progress:  20%|##        | 9/44 [02:31<02:11,  3.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  23%|##2       | 10/44 [02:32<01:41,  2.98s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.44s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.45s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.45s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.45s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.45s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.45s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:20,  2.45s/it]
Validation Progress:  25%|##5       | 11/44 [02:34<01:19,  2.42s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.08s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.08s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.07s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:05,  2.05s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.08s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.08s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.08s/it]
Validation Progress:  27%|##7       | 12/44 [02:35<01:06,  2.08s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.84s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.83s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.84s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.82s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.83s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.83s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.83s/it]
Validation Progress:  30%|##9       | 13/44 [02:36<00:56,  1.84s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.66s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.65s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.65s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.64s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.65s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.65s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.66s/it]
Validation Progress:  32%|###1      | 14/44 [02:37<00:49,  1.66s/it]
Validation Progress:  34%|###4      | 15/44 [02:38<00:44,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [02:39<00:43,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [02:38<00:44,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [02:39<00:43,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [02:39<00:43,  1.51s/it]
Validation Progress:  34%|###4      | 15/44 [02:39<00:43,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [02:39<00:44,  1.52s/it]
Validation Progress:  34%|###4      | 15/44 [02:39<00:44,  1.52s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.44s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.43s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.43s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.44s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.43s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.44s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.44s/it]
Validation Progress:  36%|###6      | 16/44 [02:40<00:40,  1.44s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:41<00:37,  1.40s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:42<00:35,  1.36s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:44<00:33,  1.35s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.32s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:45<00:31,  1.31s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:46<00:29,  1.28s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:47<00:27,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:49<00:26,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:50<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:51<00:25,  1.32s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:53<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [02:54<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:55<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:57<00:19,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:58<00:18,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.36s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.36s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.35s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.36s/it]
Validation Progress:  70%|#######   | 31/44 [02:59<00:17,  1.36s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  73%|#######2  | 32/44 [03:01<00:16,  1.35s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [03:02<00:14,  1.32s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [03:03<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [03:04<00:11,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [03:06<00:10,  1.30s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [03:07<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [03:09<00:08,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [03:10<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [03:11<00:05,  1.35s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  93%|#########3| 41/44 [03:12<00:03,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [03:14<00:02,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [03:15<00:01,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.46s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.46s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.46s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.46s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.46s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.46s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.47s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [03:16<00:00,  4.47s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.46s/it]
100%|##########| 1/1 [00:03<00:00,  3.46s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|          | 1/157 [00:01<03:05,  1.19s/it]
  2%|1         | 3/157 [00:01<00:55,  2.80it/s]
  3%|3         | 5/157 [00:01<00:31,  4.88it/s]
  4%|4         | 7/157 [00:01<00:21,  6.98it/s]
  6%|5         | 9/157 [00:01<00:16,  8.94it/s]
  7%|7         | 11/157 [00:01<00:13, 10.67it/s]
  8%|8         | 13/157 [00:01<00:12, 11.93it/s]
 10%|9         | 15/157 [00:02<00:10, 12.98it/s]
 11%|#         | 17/157 [00:02<00:10, 13.99it/s]
 12%|#2        | 19/157 [00:02<00:09, 14.71it/s]
 13%|#3        | 21/157 [00:02<00:09, 14.66it/s]
 15%|#4        | 23/157 [00:02<00:08, 15.21it/s]
 16%|#5        | 25/157 [00:02<00:08, 15.59it/s]
 17%|#7        | 27/157 [00:02<00:08, 15.97it/s]
 18%|#8        | 29/157 [00:02<00:07, 16.23it/s]
 20%|#9        | 31/157 [00:03<00:07, 16.40it/s]
 21%|##1       | 33/157 [00:03<00:07, 16.37it/s]
 22%|##2       | 35/157 [00:03<00:07, 16.47it/s]
 24%|##3       | 37/157 [00:03<00:07, 15.72it/s]
 25%|##4       | 39/157 [00:03<00:07, 16.06it/s]
 26%|##6       | 41/157 [00:03<00:07, 16.32it/s]
 27%|##7       | 43/157 [00:03<00:06, 16.44it/s]
 29%|##8       | 45/157 [00:03<00:06, 16.58it/s]
 30%|##9       | 47/157 [00:04<00:06, 16.70it/s]
 31%|###1      | 49/157 [00:04<00:06, 16.80it/s]
 32%|###2      | 51/157 [00:04<00:06, 16.77it/s]
 34%|###3      | 53/157 [00:04<00:06, 16.79it/s]
 35%|###5      | 55/157 [00:04<00:06, 16.58it/s]
 36%|###6      | 57/157 [00:04<00:06, 16.47it/s]
 38%|###7      | 59/157 [00:04<00:05, 16.48it/s]
 39%|###8      | 61/157 [00:04<00:05, 16.34it/s]
 40%|####      | 63/157 [00:04<00:05, 16.49it/s]
 41%|####1     | 65/157 [00:05<00:05, 16.45it/s]
 43%|####2     | 67/157 [00:05<00:05, 16.45it/s]
 44%|####3     | 69/157 [00:05<00:05, 15.17it/s]
 45%|####5     | 71/157 [00:05<00:05, 15.38it/s]
 46%|####6     | 73/157 [00:05<00:05, 15.61it/s]
 48%|####7     | 75/157 [00:05<00:05, 15.87it/s]
 49%|####9     | 77/157 [00:05<00:05, 15.87it/s]
 50%|#####     | 79/157 [00:05<00:04, 15.96it/s]
 52%|#####1    | 81/157 [00:06<00:04, 16.15it/s]
 53%|#####2    | 83/157 [00:06<00:04, 16.34it/s]
 54%|#####4    | 85/157 [00:06<00:04, 16.50it/s]
 55%|#####5    | 87/157 [00:06<00:04, 16.60it/s]
 57%|#####6    | 89/157 [00:06<00:04, 16.63it/s]
 58%|#####7    | 91/157 [00:06<00:03, 16.66it/s]
 59%|#####9    | 93/157 [00:06<00:03, 16.44it/s]
 61%|######    | 95/157 [00:06<00:03, 16.50it/s]
 62%|######1   | 97/157 [00:07<00:03, 16.56it/s]
 63%|######3   | 99/157 [00:07<00:03, 16.64it/s]
 64%|######4   | 101/157 [00:07<00:03, 16.67it/s]
 66%|######5   | 103/157 [00:07<00:03, 16.65it/s]
 67%|######6   | 105/157 [00:07<00:03, 16.65it/s]
 68%|######8   | 107/157 [00:07<00:03, 16.62it/s]
 69%|######9   | 109/157 [00:07<00:02, 16.51it/s]
 71%|#######   | 111/157 [00:07<00:02, 16.58it/s]
 72%|#######1  | 113/157 [00:08<00:02, 16.52it/s]
 73%|#######3  | 115/157 [00:08<00:02, 16.59it/s]
 75%|#######4  | 117/157 [00:08<00:02, 16.61it/s]
 76%|#######5  | 119/157 [00:08<00:02, 16.67it/s]
 77%|#######7  | 121/157 [00:08<00:02, 16.64it/s]
 78%|#######8  | 123/157 [00:08<00:02, 16.46it/s]
 80%|#######9  | 125/157 [00:08<00:01, 16.28it/s]
 81%|########  | 127/157 [00:08<00:01, 16.21it/s]
 82%|########2 | 129/157 [00:09<00:01, 16.42it/s]
 83%|########3 | 131/157 [00:09<00:01, 16.59it/s]
 85%|########4 | 133/157 [00:09<00:01, 13.71it/s]
 86%|########5 | 135/157 [00:09<00:01, 14.16it/s]
 87%|########7 | 137/157 [00:09<00:01, 14.88it/s]
 89%|########8 | 139/157 [00:09<00:01, 15.42it/s]
 90%|########9 | 141/157 [00:09<00:01, 15.58it/s]
 91%|#########1| 143/157 [00:09<00:00, 15.94it/s]
 92%|#########2| 145/157 [00:10<00:00, 16.12it/s]
 94%|#########3| 147/157 [00:10<00:00, 16.20it/s]
 95%|#########4| 149/157 [00:10<00:00, 16.41it/s]
 96%|#########6| 151/157 [00:10<00:00, 16.57it/s]
 97%|#########7| 153/157 [00:10<00:00, 16.66it/s]
 99%|#########8| 155/157 [00:10<00:00, 16.75it/s]
100%|##########| 157/157 [00:11<00:00,  6.16it/s]
100%|##########| 157/157 [00:11<00:00, 13.70it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:50, 15.56it/s]
  1%|          | 4/781 [00:00<00:50, 15.36it/s]
  1%|          | 6/781 [00:00<00:50, 15.28it/s]
  1%|1         | 8/781 [00:00<00:50, 15.45it/s]
  1%|1         | 10/781 [00:00<00:49, 15.53it/s]
  2%|1         | 12/781 [00:00<00:49, 15.49it/s]
  2%|1         | 14/781 [00:00<00:50, 15.31it/s]
  2%|2         | 16/781 [00:01<00:49, 15.45it/s]
  2%|2         | 18/781 [00:01<00:48, 15.59it/s]
  3%|2         | 20/781 [00:01<00:48, 15.68it/s]
  3%|2         | 22/781 [00:01<00:48, 15.73it/s]
  3%|3         | 24/781 [00:01<00:48, 15.72it/s]
  3%|3         | 26/781 [00:01<00:48, 15.72it/s]
  4%|3         | 28/781 [00:01<00:47, 15.85it/s]
  4%|3         | 30/781 [00:01<00:47, 15.95it/s]
  4%|4         | 32/781 [00:02<00:46, 15.98it/s]
  4%|4         | 34/781 [00:02<00:46, 15.91it/s]
  5%|4         | 36/781 [00:02<00:46, 15.96it/s]
  5%|4         | 38/781 [00:02<00:46, 15.95it/s]
  5%|5         | 40/781 [00:02<00:47, 15.65it/s]
  5%|5         | 42/781 [00:02<00:46, 15.75it/s]
  6%|5         | 44/781 [00:02<00:46, 15.92it/s]
  6%|5         | 46/781 [00:02<00:45, 16.12it/s]
  6%|6         | 48/781 [00:03<00:46, 15.86it/s]
  6%|6         | 50/781 [00:03<00:45, 16.17it/s]
  7%|6         | 52/781 [00:03<00:45, 16.16it/s]
  7%|6         | 54/781 [00:03<00:44, 16.31it/s]
  7%|7         | 56/781 [00:03<00:44, 16.45it/s]
  7%|7         | 58/781 [00:03<00:43, 16.49it/s]
  8%|7         | 60/781 [00:03<00:44, 16.29it/s]
  8%|7         | 62/781 [00:03<00:43, 16.37it/s]
  8%|8         | 64/781 [00:04<00:43, 16.34it/s]
  8%|8         | 66/781 [00:04<00:43, 16.39it/s]
  9%|8         | 68/781 [00:04<00:43, 16.49it/s]
  9%|8         | 70/781 [00:04<00:42, 16.67it/s]
  9%|9         | 72/781 [00:04<00:43, 16.48it/s]
  9%|9         | 74/781 [00:04<00:42, 16.68it/s]
 10%|9         | 76/781 [00:04<00:42, 16.78it/s]
 10%|9         | 78/781 [00:04<00:41, 16.84it/s]
 10%|#         | 80/781 [00:04<00:41, 16.78it/s]
 10%|#         | 82/781 [00:05<00:41, 16.86it/s]
 11%|#         | 84/781 [00:05<00:41, 16.95it/s]
 11%|#1        | 86/781 [00:05<00:40, 16.99it/s]
 11%|#1        | 88/781 [00:05<00:40, 16.96it/s]
 12%|#1        | 90/781 [00:05<00:40, 16.94it/s]
 12%|#1        | 92/781 [00:05<00:40, 16.95it/s]
 12%|#2        | 94/781 [00:05<00:41, 16.64it/s]
 12%|#2        | 96/781 [00:05<00:40, 16.72it/s]
 13%|#2        | 98/781 [00:06<00:40, 16.81it/s]
 13%|#2        | 100/781 [00:06<00:40, 16.87it/s]
 13%|#3        | 102/781 [00:06<00:40, 16.80it/s]
 13%|#3        | 104/781 [00:06<00:40, 16.63it/s]
 14%|#3        | 106/781 [00:06<00:40, 16.66it/s]
 14%|#3        | 108/781 [00:06<00:40, 16.69it/s]
 14%|#4        | 110/781 [00:06<00:40, 16.64it/s]
 14%|#4        | 112/781 [00:06<00:40, 16.67it/s]
 15%|#4        | 114/781 [00:07<00:39, 16.75it/s]
 15%|#4        | 116/781 [00:07<00:39, 16.70it/s]
 15%|#5        | 118/781 [00:07<00:41, 16.16it/s]
 15%|#5        | 120/781 [00:07<00:40, 16.30it/s]
 16%|#5        | 122/781 [00:07<00:40, 16.41it/s]
 16%|#5        | 124/781 [00:07<00:39, 16.57it/s]
 16%|#6        | 126/781 [00:07<00:39, 16.68it/s]
 16%|#6        | 128/781 [00:07<00:38, 16.75it/s]
 17%|#6        | 130/781 [00:07<00:38, 16.78it/s]
 17%|#6        | 132/781 [00:08<00:38, 16.81it/s]
 17%|#7        | 134/781 [00:08<00:38, 16.82it/s]
 17%|#7        | 136/781 [00:08<00:39, 16.27it/s]
 18%|#7        | 138/781 [00:08<00:39, 16.39it/s]
 18%|#7        | 140/781 [00:08<00:39, 16.22it/s]
 18%|#8        | 142/781 [00:08<00:39, 16.33it/s]
 18%|#8        | 144/781 [00:08<00:39, 16.26it/s]
 19%|#8        | 146/781 [00:08<00:38, 16.43it/s]
 19%|#8        | 148/781 [00:09<00:38, 16.39it/s]
 19%|#9        | 150/781 [00:09<00:38, 16.50it/s]
 19%|#9        | 152/781 [00:09<00:38, 16.25it/s]
 20%|#9        | 154/781 [00:09<00:38, 16.43it/s]
 20%|#9        | 156/781 [00:09<00:38, 16.31it/s]
 20%|##        | 158/781 [00:09<00:37, 16.40it/s]
 20%|##        | 160/781 [00:09<00:37, 16.43it/s]
 21%|##        | 162/781 [00:09<00:37, 16.45it/s]
 21%|##        | 164/781 [00:10<00:37, 16.47it/s]
 21%|##1       | 166/781 [00:10<00:37, 16.43it/s]
 22%|##1       | 168/781 [00:10<00:37, 16.49it/s]
 22%|##1       | 170/781 [00:10<00:36, 16.63it/s]
 22%|##2       | 172/781 [00:10<00:36, 16.65it/s]
 22%|##2       | 174/781 [00:10<00:36, 16.67it/s]
 23%|##2       | 176/781 [00:10<00:36, 16.52it/s]
 23%|##2       | 178/781 [00:10<00:36, 16.66it/s]
 23%|##3       | 180/781 [00:11<00:36, 16.67it/s]
 23%|##3       | 182/781 [00:11<00:35, 16.68it/s]
 24%|##3       | 184/781 [00:11<00:35, 16.62it/s]
 24%|##3       | 186/781 [00:11<00:35, 16.74it/s]
 24%|##4       | 188/781 [00:11<00:35, 16.69it/s]
 24%|##4       | 190/781 [00:11<00:35, 16.66it/s]
 25%|##4       | 192/781 [00:11<00:35, 16.77it/s]
 25%|##4       | 194/781 [00:11<00:34, 16.85it/s]
 25%|##5       | 196/781 [00:11<00:34, 16.76it/s]
 25%|##5       | 198/781 [00:12<00:35, 16.63it/s]
 26%|##5       | 200/781 [00:12<00:34, 16.67it/s]
 26%|##5       | 202/781 [00:12<00:34, 16.64it/s]
 26%|##6       | 204/781 [00:12<00:34, 16.81it/s]
 26%|##6       | 206/781 [00:12<00:34, 16.90it/s]
 27%|##6       | 208/781 [00:12<00:33, 16.87it/s]
 27%|##6       | 210/781 [00:12<00:33, 16.96it/s]
 27%|##7       | 212/781 [00:12<00:33, 16.98it/s]
 27%|##7       | 214/781 [00:13<00:33, 16.72it/s]
 28%|##7       | 216/781 [00:13<00:33, 16.68it/s]
 28%|##7       | 218/781 [00:13<00:33, 16.65it/s]
 28%|##8       | 220/781 [00:13<00:33, 16.65it/s]
 28%|##8       | 222/781 [00:13<00:33, 16.73it/s]
 29%|##8       | 224/781 [00:13<00:33, 16.85it/s]
 29%|##8       | 226/781 [00:13<00:32, 16.86it/s]
 29%|##9       | 228/781 [00:13<00:32, 16.89it/s]
 29%|##9       | 230/781 [00:13<00:32, 16.89it/s]
 30%|##9       | 232/781 [00:14<00:32, 16.82it/s]
 30%|##9       | 234/781 [00:14<00:32, 16.79it/s]
 30%|###       | 236/781 [00:14<00:32, 16.88it/s]
 30%|###       | 238/781 [00:14<00:32, 16.94it/s]
 31%|###       | 240/781 [00:14<00:31, 16.95it/s]
 31%|###       | 242/781 [00:14<00:31, 16.93it/s]
 31%|###1      | 244/781 [00:14<00:32, 16.74it/s]
 31%|###1      | 246/781 [00:14<00:32, 16.63it/s]
 32%|###1      | 248/781 [00:15<00:32, 16.58it/s]
 32%|###2      | 250/781 [00:15<00:32, 16.57it/s]
 32%|###2      | 252/781 [00:15<00:31, 16.54it/s]
 33%|###2      | 254/781 [00:15<00:31, 16.54it/s]
 33%|###2      | 256/781 [00:15<00:31, 16.47it/s]
 33%|###3      | 258/781 [00:15<00:31, 16.47it/s]
 33%|###3      | 260/781 [00:15<00:31, 16.48it/s]
 34%|###3      | 262/781 [00:16<00:42, 12.30it/s]
 34%|###3      | 264/781 [00:16<00:38, 13.28it/s]
 34%|###4      | 266/781 [00:16<00:36, 14.01it/s]
 34%|###4      | 268/781 [00:16<00:35, 14.58it/s]
 35%|###4      | 270/781 [00:16<00:33, 15.12it/s]
 35%|###4      | 272/781 [00:16<00:32, 15.54it/s]
 35%|###5      | 274/781 [00:16<00:31, 15.88it/s]
 35%|###5      | 276/781 [00:16<00:31, 16.19it/s]
 36%|###5      | 278/781 [00:17<00:30, 16.35it/s]
 36%|###5      | 280/781 [00:17<00:30, 16.58it/s]
 36%|###6      | 282/781 [00:17<00:29, 16.67it/s]
 36%|###6      | 284/781 [00:17<00:29, 16.69it/s]
 37%|###6      | 286/781 [00:17<00:29, 16.72it/s]
 37%|###6      | 288/781 [00:17<00:29, 16.62it/s]
 37%|###7      | 290/781 [00:17<00:29, 16.58it/s]
 37%|###7      | 292/781 [00:17<00:29, 16.68it/s]
 38%|###7      | 294/781 [00:17<00:29, 16.79it/s]
 38%|###7      | 296/781 [00:18<00:28, 16.90it/s]
 38%|###8      | 298/781 [00:18<00:28, 16.95it/s]
 38%|###8      | 300/781 [00:18<00:29, 16.55it/s]
 39%|###8      | 302/781 [00:18<00:29, 16.51it/s]
 39%|###8      | 304/781 [00:18<00:28, 16.56it/s]
 39%|###9      | 306/781 [00:18<00:28, 16.64it/s]
 39%|###9      | 308/781 [00:18<00:28, 16.76it/s]
 40%|###9      | 310/781 [00:18<00:27, 16.86it/s]
 40%|###9      | 312/781 [00:19<00:27, 16.84it/s]
 40%|####      | 314/781 [00:19<00:27, 16.75it/s]
 40%|####      | 316/781 [00:19<00:27, 16.70it/s]
 41%|####      | 318/781 [00:19<00:27, 16.63it/s]
 41%|####      | 320/781 [00:19<00:27, 16.58it/s]
 41%|####1     | 322/781 [00:19<00:27, 16.66it/s]
 41%|####1     | 324/781 [00:19<00:28, 16.30it/s]
 42%|####1     | 326/781 [00:19<00:28, 16.25it/s]
 42%|####1     | 328/781 [00:20<00:27, 16.37it/s]
 42%|####2     | 330/781 [00:20<00:27, 16.18it/s]
 43%|####2     | 332/781 [00:20<00:27, 16.37it/s]
 43%|####2     | 334/781 [00:20<00:26, 16.57it/s]
 43%|####3     | 336/781 [00:20<00:26, 16.61it/s]
 43%|####3     | 338/781 [00:20<00:26, 16.49it/s]
 44%|####3     | 340/781 [00:20<00:27, 16.29it/s]
 44%|####3     | 342/781 [00:20<00:27, 16.18it/s]
 44%|####4     | 344/781 [00:21<00:27, 16.02it/s]
 44%|####4     | 346/781 [00:21<00:27, 16.07it/s]
 45%|####4     | 348/781 [00:21<00:27, 15.93it/s]
 45%|####4     | 350/781 [00:21<00:26, 16.03it/s]
 45%|####5     | 352/781 [00:21<00:27, 15.86it/s]
 45%|####5     | 354/781 [00:21<00:26, 15.88it/s]
 46%|####5     | 356/781 [00:21<00:26, 15.93it/s]
 46%|####5     | 358/781 [00:21<00:26, 15.87it/s]
 46%|####6     | 360/781 [00:22<00:26, 15.85it/s]
 46%|####6     | 362/781 [00:22<00:26, 15.97it/s]
 47%|####6     | 364/781 [00:22<00:25, 16.06it/s]
 47%|####6     | 366/781 [00:22<00:25, 16.06it/s]
 47%|####7     | 368/781 [00:22<00:25, 16.10it/s]
 47%|####7     | 370/781 [00:22<00:25, 16.06it/s]
 48%|####7     | 372/781 [00:22<00:25, 16.15it/s]
 48%|####7     | 374/781 [00:22<00:25, 16.16it/s]
 48%|####8     | 376/781 [00:23<00:25, 16.19it/s]
 48%|####8     | 378/781 [00:23<00:24, 16.36it/s]
 49%|####8     | 380/781 [00:23<00:24, 16.38it/s]
 49%|####8     | 382/781 [00:23<00:24, 16.31it/s]
 49%|####9     | 384/781 [00:23<00:24, 16.21it/s]
 49%|####9     | 386/781 [00:23<00:24, 16.37it/s]
 50%|####9     | 388/781 [00:23<00:23, 16.52it/s]
 50%|####9     | 390/781 [00:23<00:23, 16.54it/s]
 50%|#####     | 392/781 [00:23<00:23, 16.69it/s]
 50%|#####     | 394/781 [00:24<00:23, 16.82it/s]
 51%|#####     | 396/781 [00:24<00:22, 16.92it/s]
 51%|#####     | 398/781 [00:24<00:22, 16.98it/s]
 51%|#####1    | 400/781 [00:24<00:22, 16.94it/s]
 51%|#####1    | 402/781 [00:24<00:22, 16.88it/s]
 52%|#####1    | 404/781 [00:24<00:22, 16.83it/s]
 52%|#####1    | 406/781 [00:24<00:22, 16.62it/s]
 52%|#####2    | 408/781 [00:24<00:22, 16.58it/s]
 52%|#####2    | 410/781 [00:25<00:22, 16.61it/s]
 53%|#####2    | 412/781 [00:25<00:22, 16.68it/s]
 53%|#####3    | 414/781 [00:25<00:21, 16.73it/s]
 53%|#####3    | 416/781 [00:25<00:21, 16.67it/s]
 54%|#####3    | 418/781 [00:25<00:21, 16.78it/s]
 54%|#####3    | 420/781 [00:25<00:21, 16.79it/s]
 54%|#####4    | 422/781 [00:25<00:21, 16.70it/s]
 54%|#####4    | 424/781 [00:25<00:21, 16.70it/s]
 55%|#####4    | 426/781 [00:25<00:21, 16.67it/s]
 55%|#####4    | 428/781 [00:26<00:21, 16.62it/s]
 55%|#####5    | 430/781 [00:26<00:20, 16.72it/s]
 55%|#####5    | 432/781 [00:26<00:20, 16.76it/s]
 56%|#####5    | 434/781 [00:26<00:20, 16.81it/s]
 56%|#####5    | 436/781 [00:26<00:20, 16.92it/s]
 56%|#####6    | 438/781 [00:26<00:20, 16.98it/s]
 56%|#####6    | 440/781 [00:26<00:20, 16.79it/s]
 57%|#####6    | 442/781 [00:26<00:20, 16.69it/s]
 57%|#####6    | 444/781 [00:27<00:20, 16.58it/s]
 57%|#####7    | 446/781 [00:27<00:20, 16.63it/s]
 57%|#####7    | 448/781 [00:27<00:20, 16.63it/s]
 58%|#####7    | 450/781 [00:27<00:19, 16.74it/s]
 58%|#####7    | 452/781 [00:27<00:19, 16.79it/s]
 58%|#####8    | 454/781 [00:27<00:19, 16.36it/s]
 58%|#####8    | 456/781 [00:28<00:30, 10.78it/s]
 59%|#####8    | 458/781 [00:28<00:27, 11.94it/s]
 59%|#####8    | 460/781 [00:28<00:24, 12.95it/s]
 59%|#####9    | 462/781 [00:28<00:23, 13.66it/s]
 59%|#####9    | 464/781 [00:28<00:22, 14.24it/s]
 60%|#####9    | 466/781 [00:28<00:21, 14.56it/s]
 60%|#####9    | 468/781 [00:28<00:20, 14.96it/s]
 60%|######    | 470/781 [00:28<00:20, 15.29it/s]
 60%|######    | 472/781 [00:29<00:19, 15.47it/s]
 61%|######    | 474/781 [00:29<00:19, 15.61it/s]
 61%|######    | 476/781 [00:29<00:19, 15.58it/s]
 61%|######1   | 478/781 [00:29<00:19, 15.18it/s]
 61%|######1   | 480/781 [00:29<00:19, 15.36it/s]
 62%|######1   | 482/781 [00:29<00:19, 15.50it/s]
 62%|######1   | 484/781 [00:29<00:19, 15.60it/s]
 62%|######2   | 486/781 [00:29<00:18, 15.57it/s]
 62%|######2   | 488/781 [00:30<00:18, 15.73it/s]
 63%|######2   | 490/781 [00:30<00:18, 15.83it/s]
 63%|######2   | 492/781 [00:30<00:18, 15.77it/s]
 63%|######3   | 494/781 [00:30<00:18, 15.81it/s]
 64%|######3   | 496/781 [00:30<00:17, 15.88it/s]
 64%|######3   | 498/781 [00:30<00:17, 15.87it/s]
 64%|######4   | 500/781 [00:30<00:17, 15.80it/s]
 64%|######4   | 502/781 [00:30<00:17, 15.87it/s]
 65%|######4   | 504/781 [00:31<00:17, 15.85it/s]
 65%|######4   | 506/781 [00:31<00:17, 15.93it/s]
 65%|######5   | 508/781 [00:31<00:17, 15.93it/s]
 65%|######5   | 510/781 [00:31<00:16, 15.98it/s]
 66%|######5   | 512/781 [00:31<00:16, 15.96it/s]
 66%|######5   | 514/781 [00:31<00:16, 15.95it/s]
 66%|######6   | 516/781 [00:31<00:16, 15.68it/s]
 66%|######6   | 518/781 [00:31<00:16, 15.65it/s]
 67%|######6   | 520/781 [00:32<00:27,  9.55it/s]
 67%|######6   | 522/781 [00:32<00:23, 10.86it/s]
 67%|######7   | 524/781 [00:32<00:21, 12.03it/s]
 67%|######7   | 526/781 [00:32<00:19, 13.01it/s]
 68%|######7   | 528/781 [00:32<00:18, 13.71it/s]
 68%|######7   | 530/781 [00:32<00:17, 14.28it/s]
 68%|######8   | 532/781 [00:33<00:16, 14.82it/s]
 68%|######8   | 534/781 [00:33<00:16, 15.19it/s]
 69%|######8   | 536/781 [00:33<00:15, 15.41it/s]
 69%|######8   | 538/781 [00:33<00:15, 15.62it/s]
 69%|######9   | 540/781 [00:33<00:15, 15.79it/s]
 69%|######9   | 542/781 [00:33<00:15, 15.83it/s]
 70%|######9   | 544/781 [00:33<00:14, 15.82it/s]
 70%|######9   | 546/781 [00:33<00:14, 15.85it/s]
 70%|#######   | 548/781 [00:34<00:14, 15.92it/s]
 70%|#######   | 550/781 [00:34<00:14, 15.98it/s]
 71%|#######   | 552/781 [00:34<00:14, 15.98it/s]
 71%|#######   | 554/781 [00:34<00:14, 15.71it/s]
 71%|#######1  | 556/781 [00:34<00:14, 15.69it/s]
 71%|#######1  | 558/781 [00:34<00:14, 15.85it/s]
 72%|#######1  | 560/781 [00:34<00:13, 15.90it/s]
 72%|#######1  | 562/781 [00:34<00:13, 16.05it/s]
 72%|#######2  | 564/781 [00:35<00:13, 16.52it/s]
 72%|#######2  | 566/781 [00:35<00:12, 16.54it/s]
 73%|#######2  | 568/781 [00:35<00:13, 16.12it/s]
 73%|#######2  | 570/781 [00:35<00:13, 15.87it/s]
 73%|#######3  | 572/781 [00:35<00:13, 15.83it/s]
 73%|#######3  | 574/781 [00:35<00:13, 15.89it/s]
 74%|#######3  | 576/781 [00:35<00:12, 15.84it/s]
 74%|#######4  | 578/781 [00:35<00:12, 15.63it/s]
 74%|#######4  | 580/781 [00:36<00:12, 15.62it/s]
 75%|#######4  | 582/781 [00:36<00:12, 15.71it/s]
 75%|#######4  | 584/781 [00:36<00:12, 15.58it/s]
 75%|#######5  | 586/781 [00:36<00:12, 15.58it/s]
 75%|#######5  | 588/781 [00:36<00:12, 15.15it/s]
 76%|#######5  | 590/781 [00:36<00:12, 15.42it/s]
 76%|#######5  | 592/781 [00:36<00:12, 15.58it/s]
 76%|#######6  | 594/781 [00:37<00:12, 15.57it/s]
 76%|#######6  | 596/781 [00:37<00:12, 15.17it/s]
 77%|#######6  | 598/781 [00:37<00:11, 15.36it/s]
 77%|#######6  | 600/781 [00:37<00:11, 15.74it/s]
 77%|#######7  | 602/781 [00:37<00:11, 16.02it/s]
 77%|#######7  | 604/781 [00:37<00:10, 16.26it/s]
 78%|#######7  | 606/781 [00:37<00:10, 16.72it/s]
 78%|#######7  | 608/781 [00:37<00:10, 16.72it/s]
 78%|#######8  | 610/781 [00:37<00:10, 16.71it/s]
 78%|#######8  | 612/781 [00:38<00:10, 16.65it/s]
 79%|#######8  | 614/781 [00:38<00:09, 16.76it/s]
 79%|#######8  | 616/781 [00:38<00:09, 16.83it/s]
 79%|#######9  | 618/781 [00:38<00:09, 16.72it/s]
 79%|#######9  | 620/781 [00:38<00:09, 16.78it/s]
 80%|#######9  | 622/781 [00:38<00:09, 16.73it/s]
 80%|#######9  | 624/781 [00:38<00:09, 16.62it/s]
 80%|########  | 626/781 [00:38<00:09, 16.63it/s]
 80%|########  | 628/781 [00:39<00:09, 16.66it/s]
 81%|########  | 630/781 [00:39<00:09, 16.68it/s]
 81%|########  | 632/781 [00:39<00:09, 16.46it/s]
 81%|########1 | 634/781 [00:39<00:08, 16.51it/s]
 81%|########1 | 636/781 [00:39<00:08, 16.55it/s]
 82%|########1 | 638/781 [00:39<00:08, 16.91it/s]
 82%|########1 | 640/781 [00:39<00:08, 16.90it/s]
 82%|########2 | 642/781 [00:39<00:08, 17.12it/s]
 82%|########2 | 644/781 [00:40<00:08, 17.05it/s]
 83%|########2 | 646/781 [00:40<00:08, 16.83it/s]
 83%|########2 | 648/781 [00:40<00:07, 16.67it/s]
 83%|########3 | 650/781 [00:40<00:08, 16.09it/s]
 83%|########3 | 652/781 [00:40<00:08, 15.75it/s]
 84%|########3 | 654/781 [00:40<00:08, 15.44it/s]
 84%|########3 | 656/781 [00:40<00:08, 15.39it/s]
 84%|########4 | 658/781 [00:40<00:08, 15.13it/s]
 85%|########4 | 660/781 [00:41<00:07, 15.38it/s]
 85%|########4 | 662/781 [00:41<00:07, 15.82it/s]
 85%|########5 | 664/781 [00:41<00:07, 15.82it/s]
 85%|########5 | 666/781 [00:41<00:07, 16.00it/s]
 86%|########5 | 668/781 [00:41<00:06, 16.30it/s]
 86%|########5 | 670/781 [00:41<00:06, 16.32it/s]
 86%|########6 | 672/781 [00:41<00:06, 16.36it/s]
 86%|########6 | 674/781 [00:41<00:06, 16.49it/s]
 87%|########6 | 676/781 [00:42<00:06, 16.50it/s]
 87%|########6 | 678/781 [00:42<00:06, 16.55it/s]
 87%|########7 | 680/781 [00:42<00:06, 16.72it/s]
 87%|########7 | 682/781 [00:42<00:05, 16.79it/s]
 88%|########7 | 684/781 [00:42<00:05, 16.91it/s]
 88%|########7 | 686/781 [00:42<00:05, 16.79it/s]
 88%|########8 | 688/781 [00:42<00:05, 16.84it/s]
 88%|########8 | 690/781 [00:42<00:05, 16.88it/s]
 89%|########8 | 692/781 [00:42<00:05, 16.87it/s]
 89%|########8 | 694/781 [00:43<00:05, 16.88it/s]
 89%|########9 | 696/781 [00:43<00:05, 16.90it/s]
 89%|########9 | 698/781 [00:43<00:04, 16.99it/s]
 90%|########9 | 700/781 [00:43<00:04, 17.00it/s]
 90%|########9 | 702/781 [00:43<00:04, 16.86it/s]
 90%|######### | 704/781 [00:43<00:04, 16.83it/s]
 90%|######### | 706/781 [00:43<00:04, 16.90it/s]
 91%|######### | 708/781 [00:43<00:04, 16.93it/s]
 91%|######### | 710/781 [00:44<00:04, 16.25it/s]
 91%|#########1| 712/781 [00:44<00:04, 16.47it/s]
 91%|#########1| 714/781 [00:44<00:04, 16.59it/s]
 92%|#########1| 716/781 [00:44<00:03, 16.68it/s]
 92%|#########1| 718/781 [00:44<00:03, 16.66it/s]
 92%|#########2| 720/781 [00:44<00:03, 16.74it/s]
 92%|#########2| 722/781 [00:44<00:03, 16.72it/s]
 93%|#########2| 724/781 [00:44<00:03, 16.71it/s]
 93%|#########2| 726/781 [00:44<00:03, 16.80it/s]
 93%|#########3| 728/781 [00:45<00:03, 16.88it/s]
 93%|#########3| 730/781 [00:45<00:03, 16.65it/s]
 94%|#########3| 732/781 [00:45<00:02, 16.53it/s]
 94%|#########3| 734/781 [00:45<00:02, 16.64it/s]
 94%|#########4| 736/781 [00:45<00:02, 16.78it/s]
 94%|#########4| 738/781 [00:45<00:02, 16.88it/s]
 95%|#########4| 740/781 [00:45<00:02, 16.93it/s]
 95%|#########5| 742/781 [00:45<00:02, 16.89it/s]
 95%|#########5| 744/781 [00:46<00:02, 16.77it/s]
 96%|#########5| 746/781 [00:46<00:02, 16.82it/s]
 96%|#########5| 748/781 [00:46<00:01, 16.80it/s]
 96%|#########6| 750/781 [00:46<00:01, 16.70it/s]
 96%|#########6| 752/781 [00:46<00:01, 16.78it/s]
 97%|#########6| 754/781 [00:46<00:01, 16.83it/s]
 97%|#########6| 756/781 [00:46<00:01, 16.89it/s]
 97%|#########7| 758/781 [00:46<00:01, 16.89it/s]
 97%|#########7| 760/781 [00:47<00:01, 16.94it/s]
 98%|#########7| 762/781 [00:47<00:01, 16.88it/s]
 98%|#########7| 764/781 [00:47<00:01, 16.90it/s]
 98%|#########8| 766/781 [00:47<00:00, 16.84it/s]
 98%|#########8| 768/781 [00:47<00:00, 16.85it/s]
 99%|#########8| 770/781 [00:47<00:00, 16.87it/s]
 99%|#########8| 772/781 [00:47<00:00, 16.92it/s]
 99%|#########9| 774/781 [00:47<00:00, 16.79it/s]
 99%|#########9| 776/781 [00:47<00:00, 16.76it/s]
100%|#########9| 778/781 [00:48<00:00, 16.86it/s]
100%|#########9| 780/781 [00:48<00:00, 16.86it/s]
100%|##########| 781/781 [00:48<00:00, 16.18it/s]
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:58,  1.37s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:58,  1.37s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:59,  1.38s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:59,  1.38s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:59,  1.39s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:59,  1.38s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:59,  1.39s/it]
Validation Progress:   2%|2         | 1/44 [00:01<01:01,  1.42s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.21s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.22s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.22s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.22s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.23s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.22s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.23s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:51,  1.24s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.27s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:52,  1.28s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:53,  1.31s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:49,  1.23s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:49,  1.24s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:49,  1.24s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:49,  1.24s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:49,  1.24s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:49,  1.24s/it]
Validation Progress:   9%|9         | 4/44 [00:05<00:50,  1.26s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.45s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.45s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.45s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.45s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.45s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.46s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.46s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:56,  1.45s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.35s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.35s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.36s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.36s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.36s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.36s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.36s/it]
Validation Progress:  14%|#3        | 6/44 [00:08<00:51,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.29s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.30s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.30s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.30s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.30s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.30s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:48,  1.30s/it]
Validation Progress:  16%|#5        | 7/44 [00:09<00:47,  1.30s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:45,  1.26s/it]
Validation Progress:  18%|#8        | 8/44 [00:10<00:46,  1.30s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.23s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:43,  1.24s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:11<00:44,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:41,  1.22s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.24s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:12<00:42,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.24s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.23s/it]
Validation Progress:  25%|##5       | 11/44 [00:14<00:40,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.24s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.25s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:15<00:39,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:39,  1.26s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:16<00:38,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:17<00:37,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:36,  1.25s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:19<00:36,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.25s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:20<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:21<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:32,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:22<00:33,  1.27s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:24<00:32,  1.29s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.27s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:25<00:30,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  48%|####7     | 21/44 [00:26<00:28,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:27<00:27,  1.24s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:29<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:30<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [00:31<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:33<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:34<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:35<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:37<00:19,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:38<00:18,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:39<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:41<00:15,  1.33s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:42<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:43<00:12,  1.27s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:44<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:46<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  84%|########4 | 37/44 [00:47<00:09,  1.37s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [00:48<00:07,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:50<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:51<00:05,  1.35s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:53<00:04,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:54<00:02,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:55<00:01,  1.33s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.91s/it]
Validation Progress: 100%|##########| 44/44 [00:58<00:00,  1.34s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:02<00:00,  2.39s/it]
100%|##########| 1/1 [00:02<00:00,  2.39s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.54it/s]
  3%|2         | 4/157 [00:00<00:11, 13.69it/s]
  4%|3         | 6/157 [00:00<00:10, 14.36it/s]
  5%|5         | 8/157 [00:00<00:09, 15.05it/s]
  6%|6         | 10/157 [00:00<00:09, 15.37it/s]
  8%|7         | 12/157 [00:00<00:09, 15.70it/s]
  9%|8         | 14/157 [00:00<00:09, 15.77it/s]
 10%|#         | 16/157 [00:01<00:08, 15.96it/s]
 11%|#1        | 18/157 [00:01<00:08, 16.01it/s]
 13%|#2        | 20/157 [00:01<00:08, 15.79it/s]
 14%|#4        | 22/157 [00:01<00:08, 15.86it/s]
 15%|#5        | 24/157 [00:01<00:08, 15.98it/s]
 17%|#6        | 26/157 [00:01<00:08, 15.96it/s]
 18%|#7        | 28/157 [00:01<00:08, 15.81it/s]
 19%|#9        | 30/157 [00:01<00:07, 15.93it/s]
 20%|##        | 32/157 [00:02<00:07, 16.03it/s]
 22%|##1       | 34/157 [00:02<00:07, 16.12it/s]
 23%|##2       | 36/157 [00:02<00:07, 15.99it/s]
 24%|##4       | 38/157 [00:02<00:07, 15.94it/s]
 25%|##5       | 40/157 [00:02<00:07, 15.96it/s]
 27%|##6       | 42/157 [00:02<00:07, 16.07it/s]
 28%|##8       | 44/157 [00:02<00:06, 16.18it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.24it/s]
 31%|###       | 48/157 [00:03<00:06, 15.93it/s]
 32%|###1      | 50/157 [00:03<00:06, 15.91it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.05it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.14it/s]
 36%|###5      | 56/157 [00:03<00:06, 16.06it/s]
 37%|###6      | 58/157 [00:03<00:06, 16.19it/s]
 38%|###8      | 60/157 [00:03<00:05, 16.24it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.23it/s]
 41%|####      | 64/157 [00:04<00:05, 16.19it/s]
 42%|####2     | 66/157 [00:04<00:05, 16.10it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.12it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.04it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.13it/s]
 47%|####7     | 74/157 [00:04<00:05, 15.92it/s]
 48%|####8     | 76/157 [00:04<00:05, 15.68it/s]
 50%|####9     | 78/157 [00:04<00:05, 15.64it/s]
 51%|#####     | 80/157 [00:05<00:04, 15.77it/s]
 52%|#####2    | 82/157 [00:05<00:04, 15.80it/s]
 54%|#####3    | 84/157 [00:05<00:04, 15.86it/s]
 55%|#####4    | 86/157 [00:05<00:04, 15.85it/s]
 56%|#####6    | 88/157 [00:05<00:04, 15.94it/s]
 57%|#####7    | 90/157 [00:05<00:04, 16.09it/s]
 59%|#####8    | 92/157 [00:05<00:04, 16.00it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.07it/s]
 61%|######1   | 96/157 [00:06<00:03, 16.17it/s]
 62%|######2   | 98/157 [00:06<00:03, 16.24it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.04it/s]
 65%|######4   | 102/157 [00:06<00:03, 15.89it/s]
 66%|######6   | 104/157 [00:06<00:03, 15.82it/s]
 68%|######7   | 106/157 [00:06<00:03, 15.89it/s]
 69%|######8   | 108/157 [00:06<00:03, 15.96it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.02it/s]
 71%|#######1  | 112/157 [00:07<00:02, 15.81it/s]
 73%|#######2  | 114/157 [00:07<00:02, 15.94it/s]
 74%|#######3  | 116/157 [00:07<00:02, 16.06it/s]
 75%|#######5  | 118/157 [00:07<00:02, 16.15it/s]
 76%|#######6  | 120/157 [00:07<00:02, 15.75it/s]
 78%|#######7  | 122/157 [00:07<00:02, 15.82it/s]
 79%|#######8  | 124/157 [00:07<00:02, 15.95it/s]
 80%|########  | 126/157 [00:07<00:01, 16.09it/s]
 82%|########1 | 128/157 [00:08<00:01, 16.03it/s]
 83%|########2 | 130/157 [00:08<00:01, 16.01it/s]
 84%|########4 | 132/157 [00:08<00:01, 16.07it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.12it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.18it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.05it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.04it/s]
 90%|######### | 142/157 [00:08<00:00, 16.02it/s]
 92%|#########1| 144/157 [00:09<00:00, 16.05it/s]
 93%|#########2| 146/157 [00:09<00:00, 16.15it/s]
 94%|#########4| 148/157 [00:09<00:00, 16.24it/s]
 96%|#########5| 150/157 [00:09<00:00, 16.16it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.05it/s]
 98%|#########8| 154/157 [00:09<00:00, 16.18it/s]
 99%|#########9| 156/157 [00:09<00:00, 15.86it/s]
100%|##########| 157/157 [00:09<00:00, 15.99it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:45, 17.02it/s]
  1%|          | 4/781 [00:00<00:45, 16.97it/s]
  1%|          | 6/781 [00:00<00:44, 17.44it/s]
  1%|1         | 8/781 [00:00<00:43, 17.67it/s]
  1%|1         | 10/781 [00:00<00:43, 17.74it/s]
  2%|1         | 12/781 [00:00<00:42, 17.94it/s]
  2%|1         | 14/781 [00:00<00:42, 18.06it/s]
  2%|2         | 16/781 [00:00<00:42, 18.17it/s]
  2%|2         | 18/781 [00:01<00:41, 18.24it/s]
  3%|2         | 20/781 [00:01<00:41, 18.32it/s]
  3%|2         | 22/781 [00:01<00:41, 18.27it/s]
  3%|3         | 24/781 [00:01<00:41, 18.11it/s]
  3%|3         | 26/781 [00:01<00:41, 18.22it/s]
  4%|3         | 28/781 [00:01<00:41, 18.23it/s]
  4%|3         | 30/781 [00:01<00:41, 18.22it/s]
  4%|4         | 32/781 [00:01<00:40, 18.29it/s]
  4%|4         | 34/781 [00:01<00:40, 18.37it/s]
  5%|4         | 36/781 [00:01<00:40, 18.21it/s]
  5%|4         | 38/781 [00:02<00:40, 18.28it/s]
  5%|5         | 40/781 [00:02<00:40, 18.33it/s]
  5%|5         | 42/781 [00:02<00:40, 18.22it/s]
  6%|5         | 44/781 [00:02<00:40, 18.20it/s]
  6%|5         | 46/781 [00:02<00:40, 18.17it/s]
  6%|6         | 48/781 [00:02<00:39, 18.42it/s]
  6%|6         | 50/781 [00:02<00:39, 18.69it/s]
  7%|6         | 53/781 [00:02<00:38, 19.05it/s]
  7%|7         | 55/781 [00:03<00:38, 18.78it/s]
  7%|7         | 57/781 [00:03<00:39, 18.44it/s]
  8%|7         | 59/781 [00:03<00:39, 18.26it/s]
  8%|7         | 61/781 [00:03<00:39, 18.14it/s]
  8%|8         | 63/781 [00:03<00:39, 18.15it/s]
  8%|8         | 65/781 [00:03<00:39, 18.20it/s]
  9%|8         | 67/781 [00:03<00:39, 18.22it/s]
  9%|8         | 69/781 [00:03<00:39, 18.15it/s]
  9%|9         | 71/781 [00:03<00:39, 18.11it/s]
  9%|9         | 73/781 [00:04<00:39, 18.07it/s]
 10%|9         | 75/781 [00:04<00:39, 17.86it/s]
 10%|9         | 77/781 [00:04<00:39, 17.78it/s]
 10%|#         | 79/781 [00:04<00:39, 17.83it/s]
 10%|#         | 81/781 [00:04<00:39, 17.80it/s]
 11%|#         | 83/781 [00:04<00:39, 17.81it/s]
 11%|#         | 85/781 [00:04<00:38, 18.06it/s]
 11%|#1        | 87/781 [00:04<00:37, 18.32it/s]
 11%|#1        | 89/781 [00:04<00:37, 18.63it/s]
 12%|#1        | 92/781 [00:05<00:35, 19.31it/s]
 12%|#2        | 95/781 [00:05<00:35, 19.49it/s]
 13%|#2        | 98/781 [00:05<00:34, 19.76it/s]
 13%|#2        | 101/781 [00:05<00:34, 19.92it/s]
 13%|#3        | 104/781 [00:05<00:33, 20.09it/s]
 14%|#3        | 107/781 [00:05<00:33, 20.17it/s]
 14%|#4        | 110/781 [00:05<00:33, 19.78it/s]
 14%|#4        | 113/781 [00:06<00:33, 19.92it/s]
 15%|#4        | 116/781 [00:06<00:33, 20.03it/s]
 15%|#5        | 119/781 [00:06<00:32, 20.21it/s]
 16%|#5        | 122/781 [00:06<00:32, 20.23it/s]
 16%|#6        | 125/781 [00:06<00:32, 20.13it/s]
 16%|#6        | 128/781 [00:06<00:32, 20.15it/s]
 17%|#6        | 131/781 [00:06<00:32, 20.14it/s]
 17%|#7        | 134/781 [00:07<00:32, 20.03it/s]
 18%|#7        | 137/781 [00:07<00:31, 20.17it/s]
 18%|#7        | 140/781 [00:07<00:31, 20.18it/s]
 18%|#8        | 143/781 [00:07<00:31, 20.01it/s]
 19%|#8        | 146/781 [00:07<00:32, 19.63it/s]
 19%|#8        | 148/781 [00:07<00:32, 19.70it/s]
 19%|#9        | 150/781 [00:07<00:31, 19.75it/s]
 20%|#9        | 153/781 [00:08<00:31, 19.93it/s]
 20%|#9        | 156/781 [00:08<00:31, 20.06it/s]
 20%|##        | 159/781 [00:08<00:31, 19.92it/s]
 21%|##        | 162/781 [00:08<00:30, 20.14it/s]
 21%|##1       | 165/781 [00:08<00:30, 20.11it/s]
 22%|##1       | 168/781 [00:08<00:31, 19.69it/s]
 22%|##1       | 170/781 [00:08<00:31, 19.71it/s]
 22%|##2       | 173/781 [00:09<00:30, 19.96it/s]
 23%|##2       | 176/781 [00:09<00:30, 20.13it/s]
 23%|##2       | 179/781 [00:09<00:29, 20.11it/s]
 23%|##3       | 182/781 [00:09<00:29, 20.17it/s]
 24%|##3       | 185/781 [00:09<00:29, 20.29it/s]
 24%|##4       | 188/781 [00:09<00:29, 20.28it/s]
 24%|##4       | 191/781 [00:09<00:29, 20.21it/s]
 25%|##4       | 194/781 [00:10<00:28, 20.35it/s]
 25%|##5       | 197/781 [00:10<00:28, 20.25it/s]
 26%|##5       | 200/781 [00:10<00:28, 20.12it/s]
 26%|##5       | 203/781 [00:10<00:28, 20.10it/s]
 26%|##6       | 206/781 [00:10<00:28, 20.17it/s]
 27%|##6       | 209/781 [00:10<00:28, 20.17it/s]
 27%|##7       | 212/781 [00:11<00:28, 20.30it/s]
 28%|##7       | 215/781 [00:11<00:27, 20.25it/s]
 28%|##7       | 218/781 [00:11<00:28, 20.03it/s]
 28%|##8       | 221/781 [00:11<00:27, 20.11it/s]
 29%|##8       | 224/781 [00:11<00:27, 20.06it/s]
 29%|##9       | 227/781 [00:11<00:27, 20.00it/s]
 29%|##9       | 230/781 [00:11<00:27, 19.95it/s]
 30%|##9       | 233/781 [00:12<00:27, 20.11it/s]
 30%|###       | 236/781 [00:12<00:27, 20.15it/s]
 31%|###       | 239/781 [00:12<00:26, 20.19it/s]
 31%|###       | 242/781 [00:12<00:26, 20.12it/s]
 31%|###1      | 245/781 [00:12<00:27, 19.74it/s]
 32%|###1      | 248/781 [00:12<00:26, 19.97it/s]
 32%|###2      | 251/781 [00:12<00:26, 20.09it/s]
 33%|###2      | 254/781 [00:13<00:26, 20.08it/s]
 33%|###2      | 257/781 [00:13<00:25, 20.23it/s]
 33%|###3      | 260/781 [00:13<00:25, 20.34it/s]
 34%|###3      | 263/781 [00:13<00:25, 20.29it/s]
 34%|###4      | 266/781 [00:13<00:25, 20.22it/s]
 34%|###4      | 269/781 [00:13<00:25, 20.21it/s]
 35%|###4      | 272/781 [00:13<00:25, 20.06it/s]
 35%|###5      | 275/781 [00:14<00:25, 20.01it/s]
 36%|###5      | 278/781 [00:14<00:25, 19.88it/s]
 36%|###5      | 280/781 [00:14<00:25, 19.90it/s]
 36%|###6      | 282/781 [00:14<00:25, 19.92it/s]
 36%|###6      | 284/781 [00:14<00:25, 19.51it/s]
 37%|###6      | 286/781 [00:14<00:25, 19.14it/s]
 37%|###6      | 288/781 [00:14<00:25, 19.08it/s]
 37%|###7      | 290/781 [00:14<00:25, 19.10it/s]
 37%|###7      | 292/781 [00:15<00:25, 19.17it/s]
 38%|###7      | 294/781 [00:15<00:25, 18.98it/s]
 38%|###7      | 296/781 [00:15<00:25, 18.70it/s]
 38%|###8      | 298/781 [00:15<00:26, 18.41it/s]
 38%|###8      | 300/781 [00:15<00:25, 18.53it/s]
 39%|###8      | 302/781 [00:15<00:25, 18.76it/s]
 39%|###8      | 304/781 [00:15<00:25, 18.58it/s]
 39%|###9      | 306/781 [00:15<00:25, 18.41it/s]
 39%|###9      | 308/781 [00:15<00:25, 18.61it/s]
 40%|###9      | 310/781 [00:16<00:24, 18.93it/s]
 40%|###9      | 312/781 [00:16<00:24, 18.87it/s]
 40%|####      | 314/781 [00:16<00:24, 18.96it/s]
 40%|####      | 316/781 [00:16<00:24, 18.98it/s]
 41%|####      | 318/781 [00:16<00:24, 19.21it/s]
 41%|####      | 320/781 [00:16<00:23, 19.37it/s]
 41%|####1     | 323/781 [00:16<00:23, 19.81it/s]
 42%|####1     | 326/781 [00:16<00:22, 20.11it/s]
 42%|####2     | 329/781 [00:16<00:22, 20.26it/s]
 43%|####2     | 332/781 [00:17<00:22, 20.36it/s]
 43%|####2     | 335/781 [00:17<00:21, 20.34it/s]
 43%|####3     | 338/781 [00:17<00:21, 20.41it/s]
 44%|####3     | 341/781 [00:17<00:21, 20.41it/s]
 44%|####4     | 344/781 [00:17<00:21, 20.18it/s]
 44%|####4     | 347/781 [00:17<00:21, 20.24it/s]
 45%|####4     | 350/781 [00:17<00:21, 20.33it/s]
 45%|####5     | 353/781 [00:18<00:21, 20.20it/s]
 46%|####5     | 356/781 [00:18<00:21, 20.22it/s]
 46%|####5     | 359/781 [00:18<00:20, 20.29it/s]
 46%|####6     | 362/781 [00:18<00:20, 20.35it/s]
 47%|####6     | 365/781 [00:18<00:20, 20.30it/s]
 47%|####7     | 368/781 [00:18<00:20, 20.22it/s]
 48%|####7     | 371/781 [00:19<00:20, 20.25it/s]
 48%|####7     | 374/781 [00:19<00:20, 20.27it/s]
 48%|####8     | 377/781 [00:19<00:19, 20.24it/s]
 49%|####8     | 380/781 [00:19<00:19, 20.15it/s]
 49%|####9     | 383/781 [00:19<00:19, 20.20it/s]
 49%|####9     | 386/781 [00:19<00:19, 20.32it/s]
 50%|####9     | 389/781 [00:19<00:19, 20.35it/s]
 50%|#####     | 392/781 [00:20<00:19, 20.37it/s]
 51%|#####     | 395/781 [00:20<00:18, 20.37it/s]
 51%|#####     | 398/781 [00:20<00:18, 20.30it/s]
 51%|#####1    | 401/781 [00:20<00:18, 20.23it/s]
 52%|#####1    | 404/781 [00:20<00:18, 20.25it/s]
 52%|#####2    | 407/781 [00:20<00:18, 20.33it/s]
 52%|#####2    | 410/781 [00:20<00:18, 20.33it/s]
 53%|#####2    | 413/781 [00:21<00:18, 20.36it/s]
 53%|#####3    | 416/781 [00:21<00:18, 20.25it/s]
 54%|#####3    | 419/781 [00:21<00:17, 20.15it/s]
 54%|#####4    | 422/781 [00:21<00:17, 20.34it/s]
 54%|#####4    | 425/781 [00:21<00:17, 20.44it/s]
 55%|#####4    | 428/781 [00:21<00:17, 20.50it/s]
 55%|#####5    | 431/781 [00:21<00:17, 20.41it/s]
 56%|#####5    | 434/781 [00:22<00:17, 20.07it/s]
 56%|#####5    | 437/781 [00:22<00:17, 20.05it/s]
 56%|#####6    | 440/781 [00:22<00:16, 20.15it/s]
 57%|#####6    | 443/781 [00:22<00:16, 20.29it/s]
 57%|#####7    | 446/781 [00:22<00:16, 20.36it/s]
 57%|#####7    | 449/781 [00:22<00:16, 20.46it/s]
 58%|#####7    | 452/781 [00:23<00:16, 20.52it/s]
 58%|#####8    | 455/781 [00:23<00:15, 20.56it/s]
 59%|#####8    | 458/781 [00:23<00:15, 20.22it/s]
 59%|#####9    | 461/781 [00:23<00:15, 20.21it/s]
 59%|#####9    | 464/781 [00:23<00:15, 20.27it/s]
 60%|#####9    | 467/781 [00:23<00:15, 20.35it/s]
 60%|######    | 470/781 [00:23<00:15, 20.26it/s]
 61%|######    | 473/781 [00:24<00:15, 20.23it/s]
 61%|######    | 476/781 [00:24<00:15, 20.11it/s]
 61%|######1   | 479/781 [00:24<00:14, 20.20it/s]
 62%|######1   | 482/781 [00:24<00:14, 20.12it/s]
 62%|######2   | 485/781 [00:24<00:14, 20.06it/s]
 62%|######2   | 488/781 [00:24<00:14, 19.99it/s]
 63%|######2   | 491/781 [00:24<00:14, 19.98it/s]
 63%|######3   | 493/781 [00:25<00:14, 19.77it/s]
 63%|######3   | 495/781 [00:25<00:14, 19.73it/s]
 64%|######3   | 498/781 [00:25<00:14, 19.87it/s]
 64%|######4   | 500/781 [00:25<00:14, 19.77it/s]
 64%|######4   | 503/781 [00:25<00:13, 20.04it/s]
 65%|######4   | 506/781 [00:25<00:13, 20.22it/s]
 65%|######5   | 509/781 [00:25<00:13, 19.99it/s]
 66%|######5   | 512/781 [00:26<00:13, 20.20it/s]
 66%|######5   | 515/781 [00:26<00:13, 20.26it/s]
 66%|######6   | 518/781 [00:26<00:13, 19.87it/s]
 67%|######6   | 521/781 [00:26<00:12, 20.03it/s]
 67%|######7   | 524/781 [00:26<00:12, 20.20it/s]
 67%|######7   | 527/781 [00:26<00:12, 20.34it/s]
 68%|######7   | 530/781 [00:26<00:12, 20.40it/s]
 68%|######8   | 533/781 [00:27<00:12, 20.41it/s]
 69%|######8   | 536/781 [00:27<00:12, 20.33it/s]
 69%|######9   | 539/781 [00:27<00:11, 20.34it/s]
 69%|######9   | 542/781 [00:27<00:11, 20.42it/s]
 70%|######9   | 545/781 [00:27<00:11, 20.44it/s]
 70%|#######   | 548/781 [00:27<00:11, 20.27it/s]
 71%|#######   | 551/781 [00:27<00:11, 20.19it/s]
 71%|#######   | 554/781 [00:28<00:11, 20.15it/s]
 71%|#######1  | 557/781 [00:28<00:11, 19.97it/s]
 72%|#######1  | 560/781 [00:28<00:10, 20.14it/s]
 72%|#######2  | 563/781 [00:28<00:10, 20.21it/s]
 72%|#######2  | 566/781 [00:28<00:10, 20.04it/s]
 73%|#######2  | 569/781 [00:28<00:10, 20.21it/s]
 73%|#######3  | 572/781 [00:28<00:10, 20.31it/s]
 74%|#######3  | 575/781 [00:29<00:10, 20.40it/s]
 74%|#######4  | 578/781 [00:29<00:10, 20.24it/s]
 74%|#######4  | 581/781 [00:29<00:09, 20.18it/s]
 75%|#######4  | 584/781 [00:29<00:09, 20.13it/s]
 75%|#######5  | 587/781 [00:29<00:09, 20.27it/s]
 76%|#######5  | 590/781 [00:29<00:09, 20.21it/s]
 76%|#######5  | 593/781 [00:30<00:09, 20.30it/s]
 76%|#######6  | 596/781 [00:30<00:09, 20.24it/s]
 77%|#######6  | 599/781 [00:30<00:09, 20.04it/s]
 77%|#######7  | 602/781 [00:30<00:08, 20.08it/s]
 77%|#######7  | 605/781 [00:30<00:08, 20.23it/s]
 78%|#######7  | 608/781 [00:30<00:08, 20.21it/s]
 78%|#######8  | 611/781 [00:30<00:08, 20.19it/s]
 79%|#######8  | 614/781 [00:31<00:08, 20.30it/s]
 79%|#######9  | 617/781 [00:31<00:08, 19.84it/s]
 79%|#######9  | 620/781 [00:31<00:08, 20.10it/s]
 80%|#######9  | 623/781 [00:31<00:07, 20.20it/s]
 80%|########  | 626/781 [00:31<00:07, 20.28it/s]
 81%|########  | 629/781 [00:31<00:07, 20.41it/s]
 81%|########  | 632/781 [00:31<00:07, 20.38it/s]
 81%|########1 | 635/781 [00:32<00:07, 20.32it/s]
 82%|########1 | 638/781 [00:32<00:07, 20.29it/s]
 82%|########2 | 641/781 [00:32<00:06, 20.24it/s]
 82%|########2 | 644/781 [00:32<00:06, 20.39it/s]
 83%|########2 | 647/781 [00:32<00:06, 20.49it/s]
 83%|########3 | 650/781 [00:32<00:06, 20.32it/s]
 84%|########3 | 653/781 [00:32<00:06, 20.18it/s]
 84%|########3 | 656/781 [00:33<00:06, 20.30it/s]
 84%|########4 | 659/781 [00:33<00:05, 20.40it/s]
 85%|########4 | 662/781 [00:33<00:05, 20.39it/s]
 85%|########5 | 665/781 [00:33<00:05, 20.45it/s]
 86%|########5 | 668/781 [00:33<00:05, 20.47it/s]
 86%|########5 | 671/781 [00:33<00:05, 20.28it/s]
 86%|########6 | 674/781 [00:34<00:05, 20.22it/s]
 87%|########6 | 677/781 [00:34<00:05, 20.15it/s]
 87%|########7 | 680/781 [00:34<00:05, 20.11it/s]
 87%|########7 | 683/781 [00:34<00:05, 19.51it/s]
 88%|########7 | 685/781 [00:34<00:04, 19.45it/s]
 88%|########7 | 687/781 [00:34<00:04, 18.97it/s]
 88%|########8 | 689/781 [00:34<00:04, 19.13it/s]
 88%|########8 | 691/781 [00:34<00:04, 18.79it/s]
 89%|########8 | 693/781 [00:35<00:04, 18.37it/s]
 89%|########8 | 695/781 [00:35<00:04, 18.48it/s]
 89%|########9 | 697/781 [00:35<00:04, 18.73it/s]
 90%|########9 | 699/781 [00:35<00:04, 19.03it/s]
 90%|########9 | 701/781 [00:35<00:04, 19.21it/s]
 90%|######### | 703/781 [00:35<00:04, 19.01it/s]
 90%|######### | 705/781 [00:35<00:03, 19.08it/s]
 91%|######### | 707/781 [00:35<00:03, 19.31it/s]
 91%|######### | 709/781 [00:35<00:03, 19.36it/s]
 91%|#########1| 711/781 [00:35<00:03, 19.06it/s]
 91%|#########1| 713/781 [00:36<00:03, 19.21it/s]
 92%|#########1| 715/781 [00:36<00:03, 19.23it/s]
 92%|#########1| 717/781 [00:36<00:03, 18.93it/s]
 92%|#########2| 719/781 [00:36<00:03, 19.04it/s]
 92%|#########2| 721/781 [00:36<00:03, 19.27it/s]
 93%|#########2| 723/781 [00:36<00:02, 19.43it/s]
 93%|#########2| 725/781 [00:36<00:02, 19.42it/s]
 93%|#########3| 727/781 [00:36<00:02, 19.44it/s]
 93%|#########3| 729/781 [00:36<00:02, 19.54it/s]
 94%|#########3| 731/781 [00:36<00:02, 19.59it/s]
 94%|#########3| 733/781 [00:37<00:02, 18.95it/s]
 94%|#########4| 735/781 [00:37<00:02, 19.21it/s]
 94%|#########4| 737/781 [00:37<00:02, 19.25it/s]
 95%|#########4| 739/781 [00:37<00:02, 19.19it/s]
 95%|#########4| 741/781 [00:37<00:02, 19.22it/s]
 95%|#########5| 743/781 [00:37<00:01, 19.33it/s]
 95%|#########5| 745/781 [00:37<00:01, 19.44it/s]
 96%|#########5| 747/781 [00:37<00:01, 19.54it/s]
 96%|#########5| 749/781 [00:37<00:01, 19.59it/s]
 96%|#########6| 751/781 [00:38<00:01, 19.62it/s]
 97%|#########6| 754/781 [00:38<00:01, 19.90it/s]
 97%|#########6| 756/781 [00:38<00:01, 19.89it/s]
 97%|#########7| 759/781 [00:38<00:01, 20.05it/s]
 98%|#########7| 762/781 [00:38<00:00, 20.12it/s]
 98%|#########7| 765/781 [00:38<00:00, 20.21it/s]
 98%|#########8| 768/781 [00:38<00:00, 20.16it/s]
 99%|#########8| 771/781 [00:39<00:00, 20.10it/s]
 99%|#########9| 774/781 [00:39<00:00, 20.15it/s]
 99%|#########9| 777/781 [00:39<00:00, 20.29it/s]
100%|#########9| 780/781 [00:39<00:00, 20.29it/s]
100%|##########| 781/781 [00:39<00:00, 19.77it/s]






=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================







=======================================================   RESTART [11-25 15:07:00]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241124_231021-c3u43coo
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run rich-plasma-547
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/c3u43coo
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_11 0.0244 ms 100.0%
  triton_mm_9 0.0244 ms 99.9%
  triton_mm_8 0.0245 ms 99.3%
  triton_mm_12 0.0247 ms 98.7%
  triton_mm_7 0.0252 ms 96.6%
  triton_mm_4 0.0253 ms 96.5%
  triton_mm_6 0.0256 ms 95.4%
  triton_mm_3 0.0260 ms 94.0%
  triton_mm_2 0.0262 ms 93.2%
  triton_mm_15 0.0269 ms 90.5%
SingleProcess AUTOTUNE benchmarking takes 2.1934 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE addmm(3432x32768, 3432x1920, 1920x32768)
  triton_mm_2898 1.8999 ms 100.0%
  triton_mm_2897 1.9167 ms 99.1%
  bias_addmm 1.9889 ms 95.5%
  triton_mm_2899 2.0669 ms 91.9%
  triton_mm_2892 2.3007 ms 82.6%
  triton_mm_2894 2.4105 ms 78.8%
  triton_mm_2890 2.4444 ms 77.7%
  addmm 2.4644 ms 77.1%
  triton_mm_2893 2.5058 ms 75.8%
  triton_mm_2895 2.7445 ms 69.2%
SingleProcess AUTOTUNE benchmarking takes 3.2082 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2672 ms 100.0%
  mm 1.4538 ms 87.2%
  triton_mm_8758 1.5493 ms 81.8%
  triton_mm_8754 1.5546 ms 81.5%
  triton_mm_8750 1.5569 ms 81.4%
  triton_mm_8755 1.6794 ms 75.5%
  triton_mm_8751 1.6992 ms 74.6%
  triton_mm_8752 1.7006 ms 74.5%
  triton_mm_8747 2.0833 ms 60.8%
  triton_mm_8748 2.1409 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.3278 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3742 ms 100.0%
  mm 0.3817 ms 98.0%
  triton_mm_11504 0.4162 ms 89.9%
  triton_mm_11508 0.4243 ms 88.2%
  triton_mm_11505 0.4468 ms 83.7%
  triton_mm_11506 0.4571 ms 81.9%
  triton_mm_11512 0.4629 ms 80.8%
  triton_mm_11509 0.4683 ms 79.9%
  triton_mm_11501 0.5594 ms 66.9%
  triton_mm_11502 0.5791 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.1615 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3958 ms 100.0%
  mm 0.4074 ms 97.1%
  triton_mm_11523 0.4388 ms 90.2%
  triton_mm_11527 0.4529 ms 87.4%
  triton_mm_11524 0.4722 ms 83.8%
  triton_mm_11525 0.4846 ms 81.7%
  triton_mm_11531 0.4902 ms 80.7%
  triton_mm_11528 0.4957 ms 79.8%
  triton_mm_11520 0.6025 ms 65.7%
  triton_mm_11521 0.6192 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 2.0773 seconds and 0.0010 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3875 ms 100.0%
  bias_addmm 5.4108 ms 99.6%
  addmm 6.0520 ms 89.0%
  triton_mm_11489 6.6559 ms 80.9%
  triton_mm_11485 6.6852 ms 80.6%
  triton_mm_11493 6.6930 ms 80.5%
  triton_mm_11490 7.1724 ms 75.1%
  triton_mm_11486 7.2801 ms 74.0%
  triton_mm_11487 7.2817 ms 74.0%
  triton_mm_11482 8.8367 ms 61.0%
SingleProcess AUTOTUNE benchmarking takes 6.5880 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1761 ms 100.0%
  bias_addmm 0.1765 ms 99.7%
  triton_mm_11546 0.1801 ms 97.8%
  triton_mm_11542 0.1806 ms 97.5%
  addmm 0.1825 ms 96.5%
  triton_mm_11543 0.1920 ms 91.7%
  triton_mm_11547 0.1933 ms 91.1%
  triton_mm_11544 0.1970 ms 89.4%
  triton_mm_11550 0.2145 ms 82.1%
  triton_mm_11540 0.2344 ms 75.1%
SingleProcess AUTOTUNE benchmarking takes 2.2112 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:53, 122.64s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2497 ms 100.0%
  mm 1.4424 ms 86.6%
  triton_mm_8758 1.5364 ms 81.3%
  triton_mm_8754 1.5452 ms 80.9%
  triton_mm_8750 1.5539 ms 80.4%
  triton_mm_8755 1.6517 ms 75.7%
  triton_mm_8751 1.6637 ms 75.1%
  triton_mm_8752 1.6673 ms 75.0%
  triton_mm_8747 2.0444 ms 61.1%
  triton_mm_8748 2.1253 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.3200 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10617 1.2586 ms 100.0%
  mm 1.2975 ms 97.0%
  triton_mm_10618 1.5387 ms 81.8%
  triton_mm_10614 1.5448 ms 81.5%
  triton_mm_10610 1.5538 ms 81.0%
  triton_mm_10615 1.6650 ms 75.6%
  triton_mm_10611 1.6655 ms 75.6%
  triton_mm_10612 1.6852 ms 74.7%
  triton_mm_10607 2.0662 ms 60.9%
  triton_mm_10608 2.1454 ms 58.7%
SingleProcess AUTOTUNE benchmarking takes 2.2289 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11361 1.2577 ms 100.0%
  mm 1.2985 ms 96.9%
  triton_mm_11358 1.5450 ms 81.4%
  triton_mm_11354 1.5523 ms 81.0%
  triton_mm_11362 1.5692 ms 80.2%
  triton_mm_11359 1.6650 ms 75.5%
  triton_mm_11355 1.6821 ms 74.8%
  triton_mm_11356 1.6967 ms 74.1%
  triton_mm_11351 2.0657 ms 60.9%
  triton_mm_11352 2.1460 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2359 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3715 ms 100.0%
  mm 0.3839 ms 96.8%
  triton_mm_11504 0.4137 ms 89.8%
  triton_mm_11508 0.4222 ms 88.0%
  triton_mm_11505 0.4451 ms 83.5%
  triton_mm_11506 0.4539 ms 81.8%
  triton_mm_11512 0.4609 ms 80.6%
  triton_mm_11509 0.4652 ms 79.8%
  triton_mm_11501 0.5609 ms 66.2%
  triton_mm_11502 0.5758 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.1589 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3944 ms 100.0%
  mm 0.4074 ms 96.8%
  triton_mm_11523 0.4368 ms 90.3%
  triton_mm_11527 0.4482 ms 88.0%
  triton_mm_11524 0.4690 ms 84.1%
  triton_mm_11525 0.4766 ms 82.8%
  triton_mm_11531 0.4879 ms 80.8%
  triton_mm_11528 0.4912 ms 80.3%
  triton_mm_11520 0.5927 ms 66.5%
  triton_mm_11521 0.6107 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.1692 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0742 ms 100.0%
  triton_mm_8668 0.0753 ms 98.5%
  triton_mm_8659 0.0762 ms 97.4%
  triton_mm_8662 0.0768 ms 96.6%
  triton_mm_8666 0.0771 ms 96.2%
  triton_mm_8661 0.0772 ms 96.0%
  triton_mm_8654 0.0774 ms 95.9%
  triton_mm_8653 0.0776 ms 95.6%
  mm 0.0777 ms 95.4%
  triton_mm_8665 0.0780 ms 95.2%
SingleProcess AUTOTUNE benchmarking takes 1.9865 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8677 0.0318 ms 100.0%
  triton_mm_8675 0.0319 ms 99.9%
  triton_mm_8681 0.0322 ms 98.9%
  triton_mm_8680 0.0329 ms 96.7%
  triton_mm_8678 0.0330 ms 96.5%
  triton_mm_8676 0.0331 ms 96.3%
  triton_mm_8673 0.0339 ms 94.0%
  triton_mm_8671 0.0351 ms 90.6%
  triton_mm_8672 0.0352 ms 90.4%
  triton_mm_8684 0.0362 ms 88.0%
SingleProcess AUTOTUNE benchmarking takes 2.1948 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9700 ms 100.0%
  triton_mm_8702 0.9727 ms 99.7%
  addmm 1.0969 ms 88.4%
  triton_mm_8699 1.1577 ms 83.8%
  triton_mm_8695 1.1640 ms 83.3%
  triton_mm_8703 1.2171 ms 79.7%
  triton_mm_8696 1.2609 ms 76.9%
  triton_mm_8697 1.2675 ms 76.5%
  triton_mm_8700 1.2678 ms 76.5%
  triton_mm_8692 1.5707 ms 61.8%
SingleProcess AUTOTUNE benchmarking takes 2.3975 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3280 ms 100.0%
  mm 0.3391 ms 96.7%
  triton_mm_8714 0.3936 ms 83.3%
  triton_mm_8718 0.3997 ms 82.1%
  triton_mm_8722 0.4024 ms 81.5%
  triton_mm_8716 0.4279 ms 76.7%
  triton_mm_8715 0.4298 ms 76.3%
  triton_mm_8719 0.4338 ms 75.6%
  triton_mm_8711 0.5223 ms 62.8%
  triton_mm_8712 0.5410 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.1422 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2586 ms 100.0%
  mm 1.2705 ms 99.1%
  triton_mm_8777 1.4795 ms 85.1%
  triton_mm_8769 1.5938 ms 79.0%
  triton_mm_8773 1.6184 ms 77.8%
  triton_mm_8771 1.6519 ms 76.2%
  triton_mm_8770 1.6766 ms 75.1%
  triton_mm_8774 1.6928 ms 74.3%
  triton_mm_8766 2.1183 ms 59.4%
  triton_mm_8767 2.1369 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.3125 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0385 ms 100.0%
  triton_mm_11428 0.0392 ms 98.3%
  triton_mm_11424 0.0397 ms 96.9%
  triton_mm_11437 0.0405 ms 95.2%
  triton_mm_11432 0.0408 ms 94.4%
  triton_mm_11422 0.0412 ms 93.4%
  triton_mm_11425 0.0413 ms 93.2%
  triton_mm_11435 0.0420 ms 91.8%
  triton_mm_11431 0.0431 ms 89.5%
  mm 0.0446 ms 86.3%
SingleProcess AUTOTUNE benchmarking takes 1.9515 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3342 ms 100.0%
  bias_addmm 5.3491 ms 99.7%
  addmm 6.0578 ms 88.1%
  triton_mm_11489 6.5750 ms 81.1%
  triton_mm_11485 6.6940 ms 79.7%
  triton_mm_11493 6.7083 ms 79.5%
  triton_mm_11490 7.1894 ms 74.2%
  triton_mm_11487 7.1960 ms 74.1%
  triton_mm_11486 7.2014 ms 74.1%
  triton_mm_11482 8.8409 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 3.0688 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1758 ms 100.0%
  bias_addmm 0.1782 ms 98.7%
  triton_mm_11542 0.1793 ms 98.0%
  triton_mm_11546 0.1798 ms 97.8%
  addmm 0.1837 ms 95.7%
  triton_mm_11543 0.1934 ms 90.9%
  triton_mm_11547 0.1944 ms 90.4%
  triton_mm_11544 0.1953 ms 90.0%
  triton_mm_11550 0.2141 ms 82.1%
  triton_mm_11540 0.2360 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 2.1940 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:53, 122.64s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2684 ms 100.0%
  mm 1.4428 ms 87.9%
  triton_mm_8758 1.5725 ms 80.7%
  triton_mm_8750 1.5783 ms 80.4%
  triton_mm_8754 1.5928 ms 79.6%
  triton_mm_8755 1.7011 ms 74.6%
  triton_mm_8751 1.7044 ms 74.4%
  triton_mm_8752 1.7139 ms 74.0%
  triton_mm_8747 2.0921 ms 60.6%
  triton_mm_8748 2.1731 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 2.3557 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9222 1.2842 ms 100.0%
  mm 1.3060 ms 98.3%
  triton_mm_9219 1.5634 ms 82.1%
  triton_mm_9223 1.5741 ms 81.6%
  triton_mm_9215 1.5797 ms 81.3%
  triton_mm_9220 1.6864 ms 76.1%
  triton_mm_9216 1.7055 ms 75.3%
  triton_mm_9217 1.7292 ms 74.3%
  triton_mm_9212 2.1127 ms 60.8%
  triton_mm_9213 2.1722 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2371 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10338 1.2802 ms 100.0%
  mm 1.3064 ms 98.0%
  triton_mm_10331 1.5797 ms 81.0%
  triton_mm_10335 1.5806 ms 81.0%
  triton_mm_10339 1.5821 ms 80.9%
  triton_mm_10336 1.7015 ms 75.2%
  triton_mm_10332 1.7056 ms 75.1%
  triton_mm_10333 1.7294 ms 74.0%
  triton_mm_10328 2.1134 ms 60.6%
  triton_mm_10329 2.1725 ms 58.9%
SingleProcess AUTOTUNE benchmarking takes 2.2537 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10710 1.2801 ms 100.0%
  mm 1.3053 ms 98.1%
  triton_mm_10711 1.5707 ms 81.5%
  triton_mm_10707 1.5936 ms 80.3%
  triton_mm_10703 1.6351 ms 78.3%
  triton_mm_10708 1.7032 ms 75.2%
  triton_mm_10704 1.7068 ms 75.0%
  triton_mm_10705 1.7361 ms 73.7%
  triton_mm_10700 2.1122 ms 60.6%
  triton_mm_10701 2.1690 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2455 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10896 1.2845 ms 100.0%
  mm 1.3060 ms 98.4%
  triton_mm_10889 1.5795 ms 81.3%
  triton_mm_10893 1.5811 ms 81.2%
  triton_mm_10897 1.5867 ms 81.0%
  triton_mm_10890 1.7055 ms 75.3%
  triton_mm_10894 1.7060 ms 75.3%
  triton_mm_10891 1.7319 ms 74.2%
  triton_mm_10886 2.1129 ms 60.8%
  triton_mm_10887 2.1722 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2524 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3730 ms 100.0%
  mm 0.3841 ms 97.1%
  triton_mm_11504 0.4195 ms 88.9%
  triton_mm_11508 0.4235 ms 88.1%
  triton_mm_11505 0.4483 ms 83.2%
  triton_mm_11506 0.4582 ms 81.4%
  triton_mm_11512 0.4629 ms 80.6%
  triton_mm_11509 0.4666 ms 79.9%
  triton_mm_11501 0.5601 ms 66.6%
  triton_mm_11502 0.5804 ms 64.3%
SingleProcess AUTOTUNE benchmarking takes 2.1679 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3959 ms 100.0%
  mm 0.4089 ms 96.8%
  triton_mm_11523 0.4382 ms 90.4%
  triton_mm_11527 0.4496 ms 88.1%
  triton_mm_11524 0.4759 ms 83.2%
  triton_mm_11525 0.4836 ms 81.9%
  triton_mm_11531 0.4894 ms 80.9%
  triton_mm_11528 0.4958 ms 79.9%
  triton_mm_11520 0.6044 ms 65.5%
  triton_mm_11521 0.6212 ms 63.7%
SingleProcess AUTOTUNE benchmarking takes 2.1834 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8676 0.0316 ms 100.0%
  triton_mm_8675 0.0317 ms 99.7%
  triton_mm_8677 0.0317 ms 99.6%
  triton_mm_8681 0.0321 ms 98.5%
  triton_mm_8680 0.0327 ms 96.6%
  triton_mm_8678 0.0327 ms 96.5%
  triton_mm_8673 0.0345 ms 91.6%
  triton_mm_8684 0.0350 ms 90.2%
  triton_mm_8671 0.0351 ms 90.1%
  triton_mm_8672 0.0360 ms 87.7%
SingleProcess AUTOTUNE benchmarking takes 2.2377 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:53, 122.64s/it]AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3676 ms 100.0%
  triton_mm_11504 0.4096 ms 89.8%
  triton_mm_11508 0.4167 ms 88.2%
  mm 0.4341 ms 84.7%
  triton_mm_11505 0.4395 ms 83.6%
  triton_mm_11512 0.4574 ms 80.4%
  triton_mm_11506 0.4582 ms 80.2%
  triton_mm_11509 0.4600 ms 79.9%
  triton_mm_11501 0.5586 ms 65.8%
  triton_mm_11502 0.5753 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 2.1415 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3893 ms 100.0%
  mm 0.4005 ms 97.2%
  triton_mm_11523 0.4324 ms 90.0%
  triton_mm_11527 0.4445 ms 87.6%
  triton_mm_11524 0.4659 ms 83.6%
  triton_mm_11525 0.4733 ms 82.2%
  triton_mm_11528 0.4879 ms 79.8%
  triton_mm_11531 0.4880 ms 79.8%
  triton_mm_11520 0.5879 ms 66.2%
  triton_mm_11521 0.6048 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.0753 seconds and 0.0010 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0747 ms 100.0%
  triton_mm_8659 0.0756 ms 98.7%
  triton_mm_8668 0.0759 ms 98.4%
  triton_mm_8662 0.0761 ms 98.2%
  triton_mm_8653 0.0770 ms 97.0%
  triton_mm_8654 0.0771 ms 96.9%
  triton_mm_8666 0.0775 ms 96.4%
  mm 0.0777 ms 96.2%
  triton_mm_8661 0.0781 ms 95.6%
  triton_mm_8665 0.0788 ms 94.8%
SingleProcess AUTOTUNE benchmarking takes 1.9767 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8678 0.0315 ms 100.0%
  triton_mm_8680 0.0316 ms 99.9%
  triton_mm_8676 0.0316 ms 99.8%
  triton_mm_8681 0.0320 ms 98.6%
  triton_mm_8675 0.0326 ms 96.7%
  triton_mm_8677 0.0326 ms 96.7%
  triton_mm_8673 0.0345 ms 91.3%
  triton_mm_8672 0.0351 ms 89.9%
  triton_mm_8684 0.0351 ms 89.9%
  triton_mm_8671 0.0362 ms 87.2%
SingleProcess AUTOTUNE benchmarking takes 2.1994 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3239 ms 100.0%
  mm 0.3376 ms 95.9%
  triton_mm_8714 0.3880 ms 83.5%
  triton_mm_8718 0.3950 ms 82.0%
  triton_mm_8722 0.4002 ms 80.9%
  triton_mm_8715 0.4259 ms 76.0%
  triton_mm_8719 0.4283 ms 75.6%
  triton_mm_8716 0.4292 ms 75.5%
  triton_mm_8711 0.5167 ms 62.7%
  triton_mm_8712 0.5355 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 2.1368 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_9386 0.0742 ms 100.0%
  triton_mm_9391 0.0760 ms 97.7%
  triton_mm_9382 0.0761 ms 97.6%
  triton_mm_9385 0.0763 ms 97.4%
  triton_mm_9376 0.0770 ms 96.4%
  triton_mm_9384 0.0774 ms 95.9%
  triton_mm_9377 0.0777 ms 95.6%
  triton_mm_9389 0.0781 ms 95.0%
  triton_mm_9388 0.0782 ms 95.0%
  mm 0.0782 ms 94.9%
SingleProcess AUTOTUNE benchmarking takes 1.9065 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9465 0.3246 ms 100.0%
  mm 0.3344 ms 97.1%
  triton_mm_9458 0.3909 ms 83.0%
  triton_mm_9462 0.3951 ms 82.2%
  triton_mm_9466 0.4007 ms 81.0%
  triton_mm_9459 0.4266 ms 76.1%
  triton_mm_9463 0.4309 ms 75.3%
  triton_mm_9460 0.4311 ms 75.3%
  triton_mm_9455 0.5214 ms 62.3%
  triton_mm_9456 0.5358 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.0586 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11424 0.0386 ms 100.0%
  triton_mm_11423 0.0392 ms 98.4%
  triton_mm_11428 0.0397 ms 97.1%
  triton_mm_11422 0.0397 ms 97.0%
  triton_mm_11437 0.0400 ms 96.3%
  triton_mm_11432 0.0404 ms 95.4%
  triton_mm_11425 0.0406 ms 95.0%
  triton_mm_11435 0.0419 ms 92.1%
  mm 0.0442 ms 87.3%
  triton_mm_11431 0.0442 ms 87.2%
SingleProcess AUTOTUNE benchmarking takes 1.9398 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3409 ms 100.0%
  bias_addmm 5.3553 ms 99.7%
  addmm 6.0425 ms 88.4%
  triton_mm_11489 6.5829 ms 81.1%
  triton_mm_11493 6.6574 ms 80.2%
  triton_mm_11485 6.6976 ms 79.7%
  triton_mm_11490 7.0852 ms 75.4%
  triton_mm_11487 7.1479 ms 74.7%
  triton_mm_11486 7.1932 ms 74.2%
  triton_mm_11482 8.8486 ms 60.4%
SingleProcess AUTOTUNE benchmarking takes 3.0344 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1742 ms 100.0%
  bias_addmm 0.1768 ms 98.6%
  triton_mm_11542 0.1775 ms 98.1%
  triton_mm_11546 0.1786 ms 97.6%
  addmm 0.1814 ms 96.0%
  triton_mm_11543 0.1912 ms 91.1%
  triton_mm_11547 0.1917 ms 90.9%
  triton_mm_11544 0.1950 ms 89.4%
  triton_mm_11550 0.2132 ms 81.7%
  triton_mm_11540 0.2343 ms 74.4%
SingleProcess AUTOTUNE benchmarking takes 2.1858 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:53, 122.64s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2461 ms 100.0%
  mm 1.4505 ms 85.9%
  triton_mm_8754 1.5335 ms 81.3%
  triton_mm_8758 1.5345 ms 81.2%
  triton_mm_8750 1.5361 ms 81.1%
  triton_mm_8755 1.6515 ms 75.5%
  triton_mm_8751 1.6573 ms 75.2%
  triton_mm_8752 1.6576 ms 75.2%
  triton_mm_8747 2.0787 ms 59.9%
  triton_mm_8748 2.1074 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.3260 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9408 1.2552 ms 100.0%
  mm 1.2897 ms 97.3%
  triton_mm_9405 1.5269 ms 82.2%
  triton_mm_9409 1.5370 ms 81.7%
  triton_mm_9401 1.5470 ms 81.1%
  triton_mm_9406 1.6666 ms 75.3%
  triton_mm_9402 1.6691 ms 75.2%
  triton_mm_9403 1.6907 ms 74.2%
  triton_mm_9398 2.0669 ms 60.7%
  triton_mm_9399 2.1252 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2281 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_9873 1.2585 ms 100.0%
  mm 1.2913 ms 97.5%
  triton_mm_9870 1.5263 ms 82.5%
  triton_mm_9874 1.5425 ms 81.6%
  triton_mm_9866 1.5464 ms 81.4%
  triton_mm_9871 1.6674 ms 75.5%
  triton_mm_9867 1.6722 ms 75.3%
  triton_mm_9868 1.6886 ms 74.5%
  triton_mm_9863 2.0672 ms 60.9%
  triton_mm_9864 2.1240 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2348 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10617 1.2559 ms 100.0%
  mm 1.2916 ms 97.2%
  triton_mm_10614 1.5271 ms 82.2%
  triton_mm_10610 1.5475 ms 81.2%
  triton_mm_10618 1.5561 ms 80.7%
  triton_mm_10615 1.6668 ms 75.4%
  triton_mm_10611 1.6687 ms 75.3%
  triton_mm_10612 1.6894 ms 74.3%
  triton_mm_10607 2.0700 ms 60.7%
  triton_mm_10608 2.1278 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2392 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10989 1.2565 ms 100.0%
  mm 1.2816 ms 98.0%
  triton_mm_10986 1.5268 ms 82.3%
  triton_mm_10982 1.5476 ms 81.2%
  triton_mm_10990 1.5560 ms 80.8%
  triton_mm_10987 1.6672 ms 75.4%
  triton_mm_10983 1.6695 ms 75.3%
  triton_mm_10984 1.6876 ms 74.5%
  triton_mm_10979 2.0710 ms 60.7%
  triton_mm_10980 2.1265 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2352 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11361 1.2558 ms 100.0%
  mm 1.2859 ms 97.7%
  triton_mm_11358 1.5273 ms 82.2%
  triton_mm_11354 1.5475 ms 81.1%
  triton_mm_11362 1.5554 ms 80.7%
  triton_mm_11359 1.6668 ms 75.3%
  triton_mm_11355 1.6688 ms 75.3%
  triton_mm_11356 1.6895 ms 74.3%
  triton_mm_11351 2.0686 ms 60.7%
  triton_mm_11352 2.1278 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.2400 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3705 ms 100.0%
  mm 0.3809 ms 97.3%
  triton_mm_11504 0.4114 ms 90.1%
  triton_mm_11508 0.4195 ms 88.3%
  triton_mm_11505 0.4429 ms 83.6%
  triton_mm_11506 0.4569 ms 81.1%
  triton_mm_11512 0.4597 ms 80.6%
  triton_mm_11509 0.4625 ms 80.1%
  triton_mm_11501 0.5606 ms 66.1%
  triton_mm_11502 0.5724 ms 64.7%
SingleProcess AUTOTUNE benchmarking takes 2.1567 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3907 ms 100.0%
  mm 0.4031 ms 96.9%
  triton_mm_11523 0.4349 ms 89.8%
  triton_mm_11527 0.4447 ms 87.9%
  triton_mm_11524 0.4666 ms 83.7%
  triton_mm_11525 0.4754 ms 82.2%
  triton_mm_11531 0.4882 ms 80.0%
  triton_mm_11528 0.4892 ms 79.9%
  triton_mm_11520 0.5908 ms 66.1%
  triton_mm_11521 0.6066 ms 64.4%
SingleProcess AUTOTUNE benchmarking takes 2.1629 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0323 ms 100.0%
  triton_mm_8678 0.0323 ms 100.0%
  triton_mm_8680 0.0323 ms 100.0%
  triton_mm_8677 0.0324 ms 99.8%
  triton_mm_8676 0.0335 ms 96.4%
  triton_mm_8681 0.0336 ms 96.1%
  triton_mm_8673 0.0342 ms 94.4%
  triton_mm_8672 0.0357 ms 90.4%
  triton_mm_8671 0.0365 ms 88.5%
  triton_mm_8684 0.0365 ms 88.4%
SingleProcess AUTOTUNE benchmarking takes 2.2197 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:53, 122.64s/it]AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3673 ms 100.0%
  triton_mm_11504 0.4086 ms 89.9%
  triton_mm_11508 0.4169 ms 88.1%
  mm 0.4310 ms 85.2%
  triton_mm_11505 0.4423 ms 83.0%
  triton_mm_11506 0.4521 ms 81.2%
  triton_mm_11509 0.4616 ms 79.6%
  triton_mm_11512 0.4617 ms 79.6%
  triton_mm_11501 0.5530 ms 66.4%
  triton_mm_11502 0.5745 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 2.1578 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3913 ms 100.0%
  mm 0.4048 ms 96.7%
  triton_mm_11523 0.4342 ms 90.1%
  triton_mm_11527 0.4436 ms 88.2%
  triton_mm_11524 0.4647 ms 84.2%
  triton_mm_11525 0.4785 ms 81.8%
  triton_mm_11531 0.4884 ms 80.1%
  triton_mm_11528 0.4899 ms 79.9%
  triton_mm_11520 0.5889 ms 66.4%
  triton_mm_11521 0.6105 ms 64.1%
SingleProcess AUTOTUNE benchmarking takes 2.1487 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x11520)
  triton_mm_8663 0.0744 ms 100.0%
  triton_mm_8668 0.0753 ms 98.7%
  triton_mm_8659 0.0755 ms 98.5%
  triton_mm_8662 0.0772 ms 96.3%
  triton_mm_8653 0.0773 ms 96.1%
  triton_mm_8654 0.0775 ms 96.0%
  mm 0.0776 ms 95.8%
  triton_mm_8666 0.0779 ms 95.4%
  triton_mm_8661 0.0782 ms 95.1%
  triton_mm_8665 0.0783 ms 95.0%
SingleProcess AUTOTUNE benchmarking takes 1.9673 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0318 ms 100.0%
  triton_mm_8677 0.0318 ms 100.0%
  triton_mm_8678 0.0329 ms 96.5%
  triton_mm_8680 0.0330 ms 96.3%
  triton_mm_8676 0.0331 ms 96.1%
  triton_mm_8681 0.0334 ms 95.0%
  triton_mm_8673 0.0335 ms 94.8%
  triton_mm_8684 0.0362 ms 87.7%
  triton_mm_8672 0.0363 ms 87.6%
  triton_mm_8671 0.0363 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 2.1945 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9664 ms 100.0%
  bias_addmm 0.9688 ms 99.7%
  addmm 1.1018 ms 87.7%
  triton_mm_8699 1.1536 ms 83.8%
  triton_mm_8695 1.1739 ms 82.3%
  triton_mm_8703 1.2151 ms 79.5%
  triton_mm_8700 1.2657 ms 76.3%
  triton_mm_8696 1.2692 ms 76.1%
  triton_mm_8697 1.2720 ms 76.0%
  triton_mm_8692 1.5716 ms 61.5%
SingleProcess AUTOTUNE benchmarking takes 2.3872 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3252 ms 100.0%
  mm 0.3366 ms 96.6%
  triton_mm_8714 0.3932 ms 82.7%
  triton_mm_8718 0.3943 ms 82.5%
  triton_mm_8722 0.4032 ms 80.7%
  triton_mm_8715 0.4255 ms 76.4%
  triton_mm_8716 0.4328 ms 75.1%
  triton_mm_8719 0.4341 ms 74.9%
  triton_mm_8711 0.5153 ms 63.1%
  triton_mm_8712 0.5366 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.1293 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2553 ms 100.0%
  mm 1.2801 ms 98.1%
  triton_mm_8777 1.4841 ms 84.6%
  triton_mm_8769 1.5956 ms 78.7%
  triton_mm_8773 1.6116 ms 77.9%
  triton_mm_8771 1.6678 ms 75.3%
  triton_mm_8770 1.6763 ms 74.9%
  triton_mm_8774 1.6898 ms 74.3%
  triton_mm_8766 2.1156 ms 59.3%
  triton_mm_8767 2.1360 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.3309 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11424 0.0390 ms 100.0%
  triton_mm_11428 0.0392 ms 99.4%
  triton_mm_11432 0.0399 ms 97.7%
  triton_mm_11423 0.0399 ms 97.7%
  triton_mm_11437 0.0401 ms 97.2%
  triton_mm_11435 0.0412 ms 94.6%
  triton_mm_11422 0.0414 ms 94.2%
  triton_mm_11425 0.0418 ms 93.2%
  triton_mm_11431 0.0436 ms 89.5%
  mm 0.0446 ms 87.5%
SingleProcess AUTOTUNE benchmarking takes 1.9453 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3900 ms 100.0%
  bias_addmm 5.4767 ms 98.4%
  addmm 6.0451 ms 89.2%
  triton_mm_11489 6.5809 ms 81.9%
  triton_mm_11485 6.6941 ms 80.5%
  triton_mm_11493 6.7093 ms 80.3%
  triton_mm_11490 7.1645 ms 75.2%
  triton_mm_11486 7.1934 ms 74.9%
  triton_mm_11487 7.2063 ms 74.8%
  triton_mm_11482 8.9313 ms 60.3%
SingleProcess AUTOTUNE benchmarking takes 3.0388 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1759 ms 100.0%
  bias_addmm 0.1779 ms 98.9%
  triton_mm_11546 0.1789 ms 98.3%
  triton_mm_11542 0.1798 ms 97.8%
  addmm 0.1823 ms 96.5%
  triton_mm_11543 0.1908 ms 92.2%
  triton_mm_11547 0.1916 ms 91.8%
  triton_mm_11544 0.1961 ms 89.7%
  triton_mm_11550 0.2146 ms 82.0%
  triton_mm_11540 0.2341 ms 75.1%
SingleProcess AUTOTUNE benchmarking takes 2.1767 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:54, 122.65s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3936 ms 100.0%
  triton_mm_11523 0.4338 ms 90.7%
  triton_mm_11527 0.4481 ms 87.9%
  mm 0.4563 ms 86.3%
  triton_mm_11524 0.4641 ms 84.8%
  triton_mm_11525 0.4797 ms 82.1%
  triton_mm_11531 0.4890 ms 80.5%
  triton_mm_11528 0.4901 ms 80.3%
  triton_mm_11520 0.5916 ms 66.5%
  triton_mm_11521 0.6136 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.0550 seconds and 0.0008 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  bias_addmm 5.3448 ms 100.0%
  triton_mm_11492 5.3948 ms 99.1%
  addmm 6.0475 ms 88.4%
  triton_mm_11489 6.5844 ms 81.2%
  triton_mm_11485 6.6923 ms 79.9%
  triton_mm_11493 6.6963 ms 79.8%
  triton_mm_11490 7.1707 ms 74.5%
  triton_mm_11486 7.1970 ms 74.3%
  triton_mm_11487 7.2115 ms 74.1%
  triton_mm_11482 8.8391 ms 60.5%
SingleProcess AUTOTUNE benchmarking takes 3.0506 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1751 ms 100.0%
  bias_addmm 0.1777 ms 98.5%
  triton_mm_11546 0.1786 ms 98.0%
  triton_mm_11542 0.1795 ms 97.6%
  addmm 0.1829 ms 95.7%
  triton_mm_11543 0.1923 ms 91.0%
  triton_mm_11547 0.1937 ms 90.4%
  triton_mm_11544 0.1972 ms 88.8%
  triton_mm_11550 0.2136 ms 82.0%
  triton_mm_11540 0.2348 ms 74.6%
SingleProcess AUTOTUNE benchmarking takes 2.1923 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:53, 122.65s/it]AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9734 ms 100.0%
  bias_addmm 1.0825 ms 89.9%
  triton_mm_8699 1.1568 ms 84.2%
  triton_mm_8695 1.1665 ms 83.4%
  triton_mm_8703 1.2187 ms 79.9%
  addmm 1.2268 ms 79.3%
  triton_mm_8700 1.2672 ms 76.8%
  triton_mm_8696 1.2682 ms 76.8%
  triton_mm_8697 1.2864 ms 75.7%
  triton_mm_8692 1.5708 ms 62.0%
SingleProcess AUTOTUNE benchmarking takes 2.3959 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3292 ms 100.0%
  mm 0.3389 ms 97.1%
  triton_mm_8714 0.3946 ms 83.4%
  triton_mm_8718 0.4010 ms 82.1%
  triton_mm_8722 0.4036 ms 81.6%
  triton_mm_8715 0.4310 ms 76.4%
  triton_mm_8716 0.4325 ms 76.1%
  triton_mm_8719 0.4352 ms 75.6%
  triton_mm_8711 0.5246 ms 62.8%
  triton_mm_8712 0.5433 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 2.1559 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2638 ms 100.0%
  mm 1.2714 ms 99.4%
  triton_mm_8777 1.4851 ms 85.1%
  triton_mm_8769 1.6029 ms 78.8%
  triton_mm_8773 1.6320 ms 77.4%
  triton_mm_8771 1.6620 ms 76.0%
  triton_mm_8770 1.6816 ms 75.2%
  triton_mm_8774 1.6949 ms 74.6%
  triton_mm_8766 2.1233 ms 59.5%
  triton_mm_8767 2.1403 ms 59.0%
SingleProcess AUTOTUNE benchmarking takes 2.3387 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8962 1.2644 ms 100.0%
  mm 1.2711 ms 99.5%
  triton_mm_8963 1.4994 ms 84.3%
  triton_mm_8955 1.6128 ms 78.4%
  triton_mm_8959 1.6339 ms 77.4%
  triton_mm_8957 1.6756 ms 75.5%
  triton_mm_8956 1.6936 ms 74.7%
  triton_mm_8960 1.6937 ms 74.7%
  triton_mm_8952 2.1370 ms 59.2%
  triton_mm_8953 2.1575 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2458 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9772 ms 100.0%
  triton_mm_8981 0.9794 ms 99.8%
  addmm 1.1064 ms 88.3%
  triton_mm_8978 1.1651 ms 83.9%
  triton_mm_8974 1.1757 ms 83.1%
  triton_mm_8982 1.2291 ms 79.5%
  triton_mm_8979 1.2807 ms 76.3%
  triton_mm_8975 1.2822 ms 76.2%
  triton_mm_8976 1.2984 ms 75.3%
  triton_mm_8971 1.5866 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 2.3098 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9000 0.3304 ms 100.0%
  mm 0.3393 ms 97.4%
  triton_mm_8993 0.3966 ms 83.3%
  triton_mm_8997 0.4001 ms 82.6%
  triton_mm_9001 0.4041 ms 81.7%
  triton_mm_8994 0.4325 ms 76.4%
  triton_mm_8995 0.4353 ms 75.9%
  triton_mm_8998 0.4360 ms 75.8%
  triton_mm_8990 0.5276 ms 62.6%
  triton_mm_8991 0.5444 ms 60.7%
SingleProcess AUTOTUNE benchmarking takes 2.0686 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0382 ms 100.0%
  triton_mm_11424 0.0391 ms 97.6%
  triton_mm_11428 0.0397 ms 96.1%
  triton_mm_11422 0.0402 ms 95.1%
  triton_mm_11437 0.0403 ms 94.8%
  triton_mm_11435 0.0407 ms 93.8%
  triton_mm_11432 0.0408 ms 93.7%
  triton_mm_11425 0.0416 ms 91.9%
  triton_mm_11431 0.0433 ms 88.1%
  mm 0.0443 ms 86.2%
SingleProcess AUTOTUNE benchmarking takes 1.9905 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3866 ms 100.0%
  bias_addmm 5.3967 ms 99.8%
  addmm 6.0480 ms 89.1%
  triton_mm_11489 6.5925 ms 81.7%
  triton_mm_11493 6.6922 ms 80.5%
  triton_mm_11485 6.6976 ms 80.4%
  triton_mm_11490 7.1848 ms 75.0%
  triton_mm_11486 7.2090 ms 74.7%
  triton_mm_11487 7.2176 ms 74.6%
  triton_mm_11482 8.8518 ms 60.9%
SingleProcess AUTOTUNE benchmarking takes 3.0539 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1765 ms 100.0%
  bias_addmm 0.1775 ms 99.4%
  triton_mm_11542 0.1790 ms 98.6%
  triton_mm_11546 0.1805 ms 97.8%
  addmm 0.1827 ms 96.6%
  triton_mm_11543 0.1939 ms 91.0%
  triton_mm_11547 0.1942 ms 90.9%
  triton_mm_11544 0.1944 ms 90.8%
  triton_mm_11550 0.2148 ms 82.2%
  triton_mm_11540 0.2343 ms 75.3%
SingleProcess AUTOTUNE benchmarking takes 2.1986 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [02:02<1:27:58, 122.75s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.67s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.67s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.68s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.67s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.67s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.67s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:52, 52.68s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:06<36:54, 52.72s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.23s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.23s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.23s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.24s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.24s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.24s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:58, 29.24s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:07<19:59, 29.26s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.15s/it]
Validation Progress:   9%|9         | 4/44 [02:08<12:06, 18.17s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.00s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.00s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.00s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.01s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.00s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.00s/it]
Validation Progress:  11%|#1        | 5/44 [02:09<07:48, 12.01s/it]
Validation Progress:  11%|#1        | 5/44 [02:10<07:48, 12.01s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  14%|#3        | 6/44 [02:11<05:15,  8.31s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:40,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:40,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:40,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:41,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:40,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:41,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:41,  5.97s/it]
Validation Progress:  16%|#5        | 7/44 [02:12<03:41,  5.98s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:39,  4.44s/it]
Validation Progress:  18%|#8        | 8/44 [02:13<02:40,  4.45s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  20%|##        | 9/44 [02:14<02:00,  3.44s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:15<01:33,  2.76s/it]
Validation Progress:  23%|##2       | 10/44 [02:16<01:35,  2.80s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.31s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.31s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.31s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:15,  2.28s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.31s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.32s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.32s/it]
Validation Progress:  25%|##5       | 11/44 [02:17<01:16,  2.31s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:02,  1.97s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  27%|##7       | 12/44 [02:18<01:03,  1.99s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.77s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.77s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.77s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.76s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.77s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.77s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:55,  1.77s/it]
Validation Progress:  30%|##9       | 13/44 [02:19<00:54,  1.77s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:47,  1.60s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:20<00:48,  1.60s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:42,  1.48s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.48s/it]
Validation Progress:  34%|###4      | 15/44 [02:22<00:43,  1.49s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.41s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.41s/it]
Validation Progress:  36%|###6      | 16/44 [02:23<00:39,  1.42s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  39%|###8      | 17/44 [02:24<00:37,  1.39s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.35s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.35s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.35s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.35s/it]
Validation Progress:  41%|####      | 18/44 [02:25<00:34,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:27<00:33,  1.35s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:28<00:31,  1.31s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:29<00:29,  1.28s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:30<00:27,  1.25s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:32<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [02:33<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [02:34<00:25,  1.32s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  59%|#####9    | 26/44 [02:36<00:23,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:37<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:38<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:40<00:19,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:41<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:42<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:44<00:15,  1.32s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [02:45<00:14,  1.29s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  77%|#######7  | 34/44 [02:46<00:12,  1.26s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:47<00:11,  1.28s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:49<00:10,  1.30s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [02:50<00:09,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [02:51<00:07,  1.31s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:53<00:06,  1.33s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.33s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.32s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.33s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.33s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.33s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.32s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.33s/it]
Validation Progress:  91%|######### | 40/44 [02:54<00:05,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  93%|#########3| 41/44 [02:55<00:03,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  95%|#########5| 42/44 [02:57<00:02,  1.29s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [02:58<00:01,  1.30s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.17s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.07s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:59<00:00,  4.08s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.38s/it]
100%|##########| 1/1 [00:03<00:00,  3.38s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|          | 1/157 [00:01<03:12,  1.24s/it]
  2%|1         | 3/157 [00:01<00:57,  2.67it/s]
  3%|3         | 5/157 [00:01<00:33,  4.58it/s]
  4%|4         | 7/157 [00:01<00:22,  6.60it/s]
  6%|5         | 9/157 [00:01<00:17,  8.52it/s]
  7%|7         | 11/157 [00:01<00:14, 10.30it/s]
  8%|8         | 13/157 [00:02<00:12, 11.50it/s]
 10%|9         | 15/157 [00:02<00:11, 12.80it/s]
 11%|#         | 17/157 [00:02<00:10, 13.86it/s]
 12%|#2        | 19/157 [00:02<00:09, 14.69it/s]
 13%|#3        | 21/157 [00:02<00:09, 14.86it/s]
 15%|#4        | 23/157 [00:02<00:08, 15.44it/s]
 16%|#5        | 25/157 [00:02<00:08, 15.85it/s]
 17%|#7        | 27/157 [00:02<00:08, 16.05it/s]
 18%|#8        | 29/157 [00:02<00:07, 16.19it/s]
 20%|#9        | 31/157 [00:03<00:07, 16.32it/s]
 21%|##1       | 33/157 [00:03<00:07, 16.42it/s]
 22%|##2       | 35/157 [00:03<00:07, 16.48it/s]
 24%|##3       | 37/157 [00:03<00:07, 15.80it/s]
 25%|##4       | 39/157 [00:03<00:07, 15.99it/s]
 26%|##6       | 41/157 [00:03<00:07, 16.24it/s]
 27%|##7       | 43/157 [00:03<00:06, 16.43it/s]
 29%|##8       | 45/157 [00:03<00:06, 16.57it/s]
 30%|##9       | 47/157 [00:04<00:06, 16.67it/s]
 31%|###1      | 49/157 [00:04<00:06, 16.74it/s]
 32%|###2      | 51/157 [00:04<00:06, 16.79it/s]
 34%|###3      | 53/157 [00:04<00:06, 16.82it/s]
 35%|###5      | 55/157 [00:04<00:06, 16.86it/s]
 36%|###6      | 57/157 [00:04<00:05, 16.89it/s]
 38%|###7      | 59/157 [00:04<00:05, 16.90it/s]
 39%|###8      | 61/157 [00:04<00:05, 16.94it/s]
 40%|####      | 63/157 [00:05<00:05, 16.94it/s]
 41%|####1     | 65/157 [00:05<00:05, 16.94it/s]
 43%|####2     | 67/157 [00:05<00:05, 16.91it/s]
 44%|####3     | 69/157 [00:05<00:05, 15.54it/s]
 45%|####5     | 71/157 [00:05<00:05, 15.89it/s]
 46%|####6     | 73/157 [00:05<00:05, 16.16it/s]
 48%|####7     | 75/157 [00:05<00:05, 16.31it/s]
 49%|####9     | 77/157 [00:05<00:04, 16.39it/s]
 50%|#####     | 79/157 [00:06<00:04, 16.44it/s]
 52%|#####1    | 81/157 [00:06<00:04, 16.52it/s]
 53%|#####2    | 83/157 [00:06<00:04, 16.63it/s]
 54%|#####4    | 85/157 [00:06<00:04, 16.69it/s]
 55%|#####5    | 87/157 [00:06<00:04, 16.76it/s]
 57%|#####6    | 89/157 [00:06<00:04, 16.81it/s]
 58%|#####7    | 91/157 [00:06<00:03, 16.82it/s]
 59%|#####9    | 93/157 [00:06<00:03, 16.81it/s]
 61%|######    | 95/157 [00:06<00:03, 16.84it/s]
 62%|######1   | 97/157 [00:07<00:03, 16.86it/s]
 63%|######3   | 99/157 [00:07<00:03, 16.91it/s]
 64%|######4   | 101/157 [00:07<00:03, 16.94it/s]
 66%|######5   | 103/157 [00:07<00:03, 16.97it/s]
 67%|######6   | 105/157 [00:07<00:03, 16.96it/s]
 68%|######8   | 107/157 [00:07<00:02, 16.95it/s]
 69%|######9   | 109/157 [00:07<00:02, 16.98it/s]
 71%|#######   | 111/157 [00:07<00:02, 16.94it/s]
 72%|#######1  | 113/157 [00:08<00:02, 16.87it/s]
 73%|#######3  | 115/157 [00:08<00:02, 16.85it/s]
 75%|#######4  | 117/157 [00:08<00:02, 16.65it/s]
 76%|#######5  | 119/157 [00:08<00:02, 16.64it/s]
 77%|#######7  | 121/157 [00:08<00:02, 16.65it/s]
 78%|#######8  | 123/157 [00:08<00:02, 16.67it/s]
 80%|#######9  | 125/157 [00:08<00:01, 16.70it/s]
 81%|########  | 127/157 [00:08<00:01, 16.73it/s]
 82%|########2 | 129/157 [00:08<00:01, 16.78it/s]
 83%|########3 | 131/157 [00:09<00:01, 16.80it/s]
 85%|########4 | 133/157 [00:09<00:01, 14.43it/s]
 86%|########5 | 135/157 [00:09<00:01, 15.03it/s]
 87%|########7 | 137/157 [00:09<00:01, 15.54it/s]
 89%|########8 | 139/157 [00:09<00:01, 15.92it/s]
 90%|########9 | 141/157 [00:09<00:00, 16.19it/s]
 91%|#########1| 143/157 [00:09<00:00, 16.41it/s]
 92%|#########2| 145/157 [00:09<00:00, 16.55it/s]
 94%|#########3| 147/157 [00:10<00:00, 16.68it/s]
 95%|#########4| 149/157 [00:10<00:00, 16.66it/s]
 96%|#########6| 151/157 [00:10<00:00, 16.68it/s]
 97%|#########7| 153/157 [00:10<00:00, 16.67it/s]
 99%|#########8| 155/157 [00:10<00:00, 16.63it/s]
100%|##########| 157/157 [00:11<00:00,  6.32it/s]
100%|##########| 157/157 [00:11<00:00, 13.81it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:52, 14.75it/s]
  1%|          | 4/781 [00:00<00:54, 14.24it/s]
  1%|          | 6/781 [00:00<00:51, 14.96it/s]
  1%|1         | 8/781 [00:00<00:50, 15.36it/s]
  1%|1         | 10/781 [00:00<00:49, 15.44it/s]
  2%|1         | 12/781 [00:00<00:49, 15.67it/s]
  2%|1         | 14/781 [00:00<00:48, 15.85it/s]
  2%|2         | 16/781 [00:01<00:48, 15.90it/s]
  2%|2         | 18/781 [00:01<00:48, 15.88it/s]
  3%|2         | 20/781 [00:01<00:47, 15.92it/s]
  3%|2         | 22/781 [00:01<00:47, 15.93it/s]
  3%|3         | 24/781 [00:01<00:47, 15.85it/s]
  3%|3         | 26/781 [00:01<00:47, 15.80it/s]
  4%|3         | 28/781 [00:01<00:47, 15.80it/s]
  4%|3         | 30/781 [00:01<00:47, 15.80it/s]
  4%|4         | 32/781 [00:02<00:47, 15.80it/s]
  4%|4         | 34/781 [00:02<00:47, 15.80it/s]
  5%|4         | 36/781 [00:02<00:47, 15.81it/s]
  5%|4         | 38/781 [00:02<00:47, 15.79it/s]
  5%|5         | 40/781 [00:02<00:46, 15.79it/s]
  5%|5         | 42/781 [00:02<00:46, 15.86it/s]
  6%|5         | 44/781 [00:02<00:46, 15.93it/s]
  6%|5         | 46/781 [00:02<00:45, 16.00it/s]
  6%|6         | 48/781 [00:03<00:45, 15.97it/s]
  6%|6         | 50/781 [00:03<00:45, 15.93it/s]
  7%|6         | 52/781 [00:03<00:45, 15.89it/s]
  7%|6         | 54/781 [00:03<00:45, 15.94it/s]
  7%|7         | 56/781 [00:03<00:45, 16.02it/s]
  7%|7         | 58/781 [00:03<00:45, 16.06it/s]
  8%|7         | 60/781 [00:03<00:44, 16.09it/s]
  8%|7         | 62/781 [00:03<00:44, 16.10it/s]
  8%|8         | 64/781 [00:04<00:44, 16.06it/s]
  8%|8         | 66/781 [00:04<00:44, 16.11it/s]
  9%|8         | 68/781 [00:04<00:44, 16.13it/s]
  9%|8         | 70/781 [00:04<00:44, 16.13it/s]
  9%|9         | 72/781 [00:04<00:43, 16.12it/s]
  9%|9         | 74/781 [00:04<00:43, 16.13it/s]
 10%|9         | 76/781 [00:04<00:43, 16.14it/s]
 10%|9         | 78/781 [00:04<00:43, 16.15it/s]
 10%|#         | 80/781 [00:05<00:43, 16.14it/s]
 10%|#         | 82/781 [00:05<00:43, 16.14it/s]
 11%|#         | 84/781 [00:05<00:43, 16.08it/s]
 11%|#1        | 86/781 [00:05<00:43, 15.93it/s]
 11%|#1        | 88/781 [00:05<00:43, 15.84it/s]
 12%|#1        | 90/781 [00:05<00:43, 15.82it/s]
 12%|#1        | 92/781 [00:05<00:43, 15.80it/s]
 12%|#2        | 94/781 [00:05<00:43, 15.77it/s]
 12%|#2        | 96/781 [00:06<00:43, 15.76it/s]
 13%|#2        | 98/781 [00:06<00:43, 15.76it/s]
 13%|#2        | 100/781 [00:06<00:42, 15.85it/s]
 13%|#3        | 102/781 [00:06<00:42, 15.86it/s]
 13%|#3        | 104/781 [00:06<00:42, 15.82it/s]
 14%|#3        | 106/781 [00:06<00:42, 15.90it/s]
 14%|#3        | 108/781 [00:06<00:42, 15.97it/s]
 14%|#4        | 110/781 [00:06<00:41, 15.99it/s]
 14%|#4        | 112/781 [00:07<00:41, 16.04it/s]
 15%|#4        | 114/781 [00:07<00:41, 16.07it/s]
 15%|#4        | 116/781 [00:07<00:41, 16.10it/s]
 15%|#5        | 118/781 [00:07<00:41, 16.11it/s]
 15%|#5        | 120/781 [00:07<00:40, 16.15it/s]
 16%|#5        | 122/781 [00:07<00:41, 16.07it/s]
 16%|#5        | 124/781 [00:07<00:40, 16.10it/s]
 16%|#6        | 126/781 [00:07<00:40, 16.17it/s]
 16%|#6        | 128/781 [00:08<00:40, 16.26it/s]
 17%|#6        | 130/781 [00:08<00:39, 16.32it/s]
 17%|#6        | 132/781 [00:08<00:39, 16.29it/s]
 17%|#7        | 134/781 [00:08<00:39, 16.30it/s]
 17%|#7        | 136/781 [00:08<00:39, 16.23it/s]
 18%|#7        | 138/781 [00:08<00:40, 15.94it/s]
 18%|#7        | 140/781 [00:08<00:40, 16.01it/s]
 18%|#8        | 142/781 [00:08<00:39, 16.07it/s]
 18%|#8        | 144/781 [00:09<00:39, 15.97it/s]
 19%|#8        | 146/781 [00:09<00:39, 15.91it/s]
 19%|#8        | 148/781 [00:09<00:39, 15.96it/s]
 19%|#9        | 150/781 [00:09<00:39, 15.96it/s]
 19%|#9        | 152/781 [00:09<00:38, 16.23it/s]
 20%|#9        | 154/781 [00:09<00:38, 16.20it/s]
 20%|#9        | 156/781 [00:09<00:38, 16.28it/s]
 20%|##        | 158/781 [00:09<00:38, 16.26it/s]
 20%|##        | 160/781 [00:10<00:38, 16.10it/s]
 21%|##        | 162/781 [00:10<00:38, 16.18it/s]
 21%|##        | 164/781 [00:10<00:38, 16.06it/s]
 21%|##1       | 166/781 [00:10<00:38, 16.16it/s]
 22%|##1       | 168/781 [00:10<00:37, 16.21it/s]
 22%|##1       | 170/781 [00:10<00:37, 16.19it/s]
 22%|##2       | 172/781 [00:10<00:37, 16.15it/s]
 22%|##2       | 174/781 [00:10<00:37, 16.35it/s]
 23%|##2       | 176/781 [00:11<00:36, 16.51it/s]
 23%|##2       | 178/781 [00:11<00:36, 16.62it/s]
 23%|##3       | 180/781 [00:11<00:36, 16.68it/s]
 23%|##3       | 182/781 [00:11<00:36, 16.61it/s]
 24%|##3       | 184/781 [00:11<00:36, 16.47it/s]
 24%|##3       | 186/781 [00:11<00:36, 16.33it/s]
 24%|##4       | 188/781 [00:11<00:36, 16.28it/s]
 24%|##4       | 190/781 [00:11<00:36, 16.35it/s]
 25%|##4       | 192/781 [00:11<00:36, 16.29it/s]
 25%|##4       | 194/781 [00:12<00:36, 16.25it/s]
 25%|##5       | 196/781 [00:12<00:35, 16.26it/s]
 25%|##5       | 198/781 [00:12<00:36, 16.19it/s]
 26%|##5       | 200/781 [00:12<00:35, 16.22it/s]
 26%|##5       | 202/781 [00:12<00:35, 16.29it/s]
 26%|##6       | 204/781 [00:12<00:35, 16.35it/s]
 26%|##6       | 206/781 [00:12<00:34, 16.51it/s]
 27%|##6       | 208/781 [00:12<00:34, 16.50it/s]
 27%|##6       | 210/781 [00:13<00:34, 16.36it/s]
 27%|##7       | 212/781 [00:13<00:34, 16.39it/s]
 27%|##7       | 214/781 [00:13<00:34, 16.54it/s]
 28%|##7       | 216/781 [00:13<00:34, 16.41it/s]
 28%|##7       | 218/781 [00:13<00:33, 16.57it/s]
 28%|##8       | 220/781 [00:13<00:33, 16.54it/s]
 28%|##8       | 222/781 [00:13<00:34, 16.42it/s]
 29%|##8       | 224/781 [00:13<00:33, 16.44it/s]
 29%|##8       | 226/781 [00:14<00:33, 16.44it/s]
 29%|##9       | 228/781 [00:14<00:33, 16.36it/s]
 29%|##9       | 230/781 [00:14<00:33, 16.41it/s]
 30%|##9       | 232/781 [00:14<00:33, 16.34it/s]
 30%|##9       | 234/781 [00:14<00:33, 16.41it/s]
 30%|###       | 236/781 [00:14<00:33, 16.44it/s]
 30%|###       | 238/781 [00:14<00:32, 16.49it/s]
 31%|###       | 240/781 [00:14<00:32, 16.50it/s]
 31%|###       | 242/781 [00:15<00:32, 16.35it/s]
 31%|###1      | 244/781 [00:15<00:33, 16.26it/s]
 31%|###1      | 246/781 [00:15<00:32, 16.22it/s]
 32%|###1      | 248/781 [00:15<00:33, 16.09it/s]
 32%|###2      | 250/781 [00:15<00:32, 16.18it/s]
 32%|###2      | 252/781 [00:15<00:32, 16.08it/s]
 33%|###2      | 254/781 [00:15<00:32, 16.13it/s]
 33%|###2      | 256/781 [00:15<00:32, 16.16it/s]
 33%|###3      | 258/781 [00:16<00:32, 16.17it/s]
 33%|###3      | 260/781 [00:16<00:31, 16.38it/s]
 34%|###3      | 262/781 [00:16<00:46, 11.24it/s]
 34%|###3      | 264/781 [00:16<00:41, 12.31it/s]
 34%|###4      | 266/781 [00:16<00:39, 13.18it/s]
 34%|###4      | 268/781 [00:16<00:37, 13.86it/s]
 35%|###4      | 270/781 [00:16<00:35, 14.35it/s]
 35%|###4      | 272/781 [00:17<00:34, 14.60it/s]
 35%|###5      | 274/781 [00:17<00:33, 15.03it/s]
 35%|###5      | 276/781 [00:17<00:33, 15.24it/s]
 36%|###5      | 278/781 [00:17<00:32, 15.39it/s]
 36%|###5      | 280/781 [00:17<00:32, 15.60it/s]
 36%|###6      | 282/781 [00:17<00:31, 15.71it/s]
 36%|###6      | 284/781 [00:17<00:31, 15.92it/s]
 37%|###6      | 286/781 [00:17<00:30, 16.09it/s]
 37%|###6      | 288/781 [00:18<00:30, 16.20it/s]
 37%|###7      | 290/781 [00:18<00:30, 16.15it/s]
 37%|###7      | 292/781 [00:18<00:30, 16.24it/s]
 38%|###7      | 294/781 [00:18<00:30, 16.11it/s]
 38%|###7      | 296/781 [00:18<00:29, 16.33it/s]
 38%|###8      | 298/781 [00:18<00:29, 16.49it/s]
 38%|###8      | 300/781 [00:18<00:29, 16.51it/s]
 39%|###8      | 302/781 [00:18<00:29, 16.50it/s]
 39%|###8      | 304/781 [00:19<00:28, 16.52it/s]
 39%|###9      | 306/781 [00:19<00:28, 16.39it/s]
 39%|###9      | 308/781 [00:19<00:28, 16.44it/s]
 40%|###9      | 310/781 [00:19<00:28, 16.41it/s]
 40%|###9      | 312/781 [00:19<00:28, 16.31it/s]
 40%|####      | 314/781 [00:19<00:28, 16.32it/s]
 40%|####      | 316/781 [00:19<00:28, 16.26it/s]
 41%|####      | 318/781 [00:19<00:28, 16.10it/s]
 41%|####      | 320/781 [00:20<00:28, 16.12it/s]
 41%|####1     | 322/781 [00:20<00:28, 16.02it/s]
 41%|####1     | 324/781 [00:20<00:28, 15.99it/s]
 42%|####1     | 326/781 [00:20<00:28, 16.00it/s]
 42%|####1     | 328/781 [00:20<00:28, 15.96it/s]
 42%|####2     | 330/781 [00:20<00:28, 16.00it/s]
 43%|####2     | 332/781 [00:20<00:27, 16.07it/s]
 43%|####2     | 334/781 [00:20<00:27, 16.09it/s]
 43%|####3     | 336/781 [00:21<00:27, 16.13it/s]
 43%|####3     | 338/781 [00:21<00:27, 16.19it/s]
 44%|####3     | 340/781 [00:21<00:27, 16.18it/s]
 44%|####3     | 342/781 [00:21<00:27, 16.18it/s]
 44%|####4     | 344/781 [00:21<00:27, 16.18it/s]
 44%|####4     | 346/781 [00:21<00:26, 16.19it/s]
 45%|####4     | 348/781 [00:21<00:26, 16.20it/s]
 45%|####4     | 350/781 [00:21<00:26, 16.21it/s]
 45%|####5     | 352/781 [00:22<00:26, 16.21it/s]
 45%|####5     | 354/781 [00:22<00:26, 16.20it/s]
 46%|####5     | 356/781 [00:22<00:26, 16.22it/s]
 46%|####5     | 358/781 [00:22<00:26, 16.22it/s]
 46%|####6     | 360/781 [00:22<00:25, 16.20it/s]
 46%|####6     | 362/781 [00:22<00:25, 16.19it/s]
 47%|####6     | 364/781 [00:22<00:25, 16.20it/s]
 47%|####6     | 366/781 [00:22<00:25, 16.20it/s]
 47%|####7     | 368/781 [00:23<00:25, 16.21it/s]
 47%|####7     | 370/781 [00:23<00:25, 16.21it/s]
 48%|####7     | 372/781 [00:23<00:25, 16.21it/s]
 48%|####7     | 374/781 [00:23<00:25, 16.22it/s]
 48%|####8     | 376/781 [00:23<00:24, 16.21it/s]
 48%|####8     | 378/781 [00:23<00:24, 16.24it/s]
 49%|####8     | 380/781 [00:23<00:24, 16.22it/s]
 49%|####8     | 382/781 [00:23<00:24, 16.21it/s]
 49%|####9     | 384/781 [00:24<00:24, 16.19it/s]
 49%|####9     | 386/781 [00:24<00:24, 16.19it/s]
 50%|####9     | 388/781 [00:24<00:24, 16.19it/s]
 50%|####9     | 390/781 [00:24<00:24, 16.17it/s]
 50%|#####     | 392/781 [00:24<00:24, 16.18it/s]
 50%|#####     | 394/781 [00:24<00:23, 16.19it/s]
 51%|#####     | 396/781 [00:24<00:23, 16.20it/s]
 51%|#####     | 398/781 [00:24<00:23, 16.22it/s]
 51%|#####1    | 400/781 [00:24<00:23, 16.22it/s]
 51%|#####1    | 402/781 [00:25<00:23, 16.09it/s]
 52%|#####1    | 404/781 [00:25<00:23, 16.09it/s]
 52%|#####1    | 406/781 [00:25<00:23, 16.09it/s]
 52%|#####2    | 408/781 [00:25<00:23, 16.04it/s]
 52%|#####2    | 410/781 [00:25<00:23, 15.99it/s]
 53%|#####2    | 412/781 [00:25<00:23, 15.96it/s]
 53%|#####3    | 414/781 [00:25<00:23, 15.94it/s]
 53%|#####3    | 416/781 [00:25<00:22, 15.91it/s]
 54%|#####3    | 418/781 [00:26<00:22, 15.92it/s]
 54%|#####3    | 420/781 [00:26<00:22, 15.95it/s]
 54%|#####4    | 422/781 [00:26<00:22, 16.01it/s]
 54%|#####4    | 424/781 [00:26<00:22, 16.06it/s]
 55%|#####4    | 426/781 [00:26<00:22, 16.11it/s]
 55%|#####4    | 428/781 [00:26<00:22, 15.97it/s]
 55%|#####5    | 430/781 [00:26<00:21, 15.99it/s]
 55%|#####5    | 432/781 [00:26<00:21, 16.05it/s]
 56%|#####5    | 434/781 [00:27<00:21, 16.05it/s]
 56%|#####5    | 436/781 [00:27<00:21, 16.07it/s]
 56%|#####6    | 438/781 [00:27<00:21, 16.11it/s]
 56%|#####6    | 440/781 [00:27<00:21, 16.12it/s]
 57%|#####6    | 442/781 [00:27<00:20, 16.15it/s]
 57%|#####6    | 444/781 [00:27<00:20, 16.18it/s]
 57%|#####7    | 446/781 [00:27<00:20, 16.18it/s]
 57%|#####7    | 448/781 [00:27<00:20, 16.19it/s]
 58%|#####7    | 450/781 [00:28<00:20, 16.21it/s]
 58%|#####7    | 452/781 [00:28<00:20, 16.20it/s]
 58%|#####8    | 454/781 [00:28<00:20, 16.18it/s]
 58%|#####8    | 456/781 [00:28<00:20, 16.19it/s]
 59%|#####8    | 458/781 [00:28<00:19, 16.18it/s]
 59%|#####8    | 460/781 [00:28<00:19, 16.16it/s]
 59%|#####9    | 462/781 [00:28<00:19, 16.17it/s]
 59%|#####9    | 464/781 [00:28<00:19, 16.18it/s]
 60%|#####9    | 466/781 [00:29<00:19, 16.02it/s]
 60%|#####9    | 468/781 [00:29<00:19, 15.99it/s]
 60%|######    | 470/781 [00:29<00:19, 15.93it/s]
 60%|######    | 472/781 [00:29<00:19, 15.98it/s]
 61%|######    | 474/781 [00:29<00:19, 15.87it/s]
 61%|######    | 476/781 [00:29<00:19, 15.93it/s]
 61%|######1   | 478/781 [00:29<00:19, 15.89it/s]
 61%|######1   | 480/781 [00:29<00:18, 15.92it/s]
 62%|######1   | 482/781 [00:30<00:18, 15.91it/s]
 62%|######1   | 484/781 [00:30<00:18, 15.91it/s]
 62%|######2   | 486/781 [00:30<00:18, 15.86it/s]
 62%|######2   | 488/781 [00:30<00:18, 15.90it/s]
 63%|######2   | 490/781 [00:30<00:18, 15.95it/s]
 63%|######2   | 492/781 [00:30<00:18, 15.84it/s]
 63%|######3   | 494/781 [00:30<00:18, 15.81it/s]
 64%|######3   | 496/781 [00:30<00:18, 15.79it/s]
 64%|######3   | 498/781 [00:31<00:17, 15.79it/s]
 64%|######4   | 500/781 [00:31<00:17, 15.80it/s]
 64%|######4   | 502/781 [00:31<00:17, 15.84it/s]
 65%|######4   | 504/781 [00:31<00:17, 15.93it/s]
 65%|######4   | 506/781 [00:31<00:17, 16.02it/s]
 65%|######5   | 508/781 [00:31<00:17, 16.06it/s]
 65%|######5   | 510/781 [00:31<00:16, 16.11it/s]
 66%|######5   | 512/781 [00:31<00:16, 16.13it/s]
 66%|######5   | 514/781 [00:32<00:16, 16.08it/s]
 66%|######6   | 516/781 [00:32<00:16, 16.02it/s]
 66%|######6   | 518/781 [00:32<00:16, 15.92it/s]
 67%|######6   | 520/781 [00:32<00:26,  9.80it/s]
 67%|######6   | 522/781 [00:32<00:23, 11.02it/s]
 67%|######7   | 524/781 [00:33<00:21, 12.09it/s]
 67%|######7   | 526/781 [00:33<00:19, 12.98it/s]
 68%|######7   | 528/781 [00:33<00:18, 13.69it/s]
 68%|######7   | 530/781 [00:33<00:17, 14.25it/s]
 68%|######8   | 532/781 [00:33<00:16, 14.66it/s]
 68%|######8   | 534/781 [00:33<00:16, 14.98it/s]
 69%|######8   | 536/781 [00:33<00:16, 15.24it/s]
 69%|######8   | 538/781 [00:33<00:15, 15.40it/s]
 69%|######9   | 540/781 [00:34<00:15, 15.51it/s]
 69%|######9   | 542/781 [00:34<00:15, 15.58it/s]
 70%|######9   | 544/781 [00:34<00:15, 15.66it/s]
 70%|######9   | 546/781 [00:34<00:14, 15.81it/s]
 70%|#######   | 548/781 [00:34<00:14, 15.93it/s]
 70%|#######   | 550/781 [00:34<00:14, 16.01it/s]
 71%|#######   | 552/781 [00:34<00:14, 16.07it/s]
 71%|#######   | 554/781 [00:34<00:14, 16.09it/s]
 71%|#######1  | 556/781 [00:35<00:13, 16.11it/s]
 71%|#######1  | 558/781 [00:35<00:13, 16.02it/s]
 72%|#######1  | 560/781 [00:35<00:13, 15.98it/s]
 72%|#######1  | 562/781 [00:35<00:13, 15.97it/s]
 72%|#######2  | 564/781 [00:35<00:13, 16.00it/s]
 72%|#######2  | 566/781 [00:35<00:13, 16.02it/s]
 73%|#######2  | 568/781 [00:35<00:13, 16.05it/s]
 73%|#######2  | 570/781 [00:35<00:13, 16.08it/s]
 73%|#######3  | 572/781 [00:36<00:12, 16.12it/s]
 73%|#######3  | 574/781 [00:36<00:12, 16.13it/s]
 74%|#######3  | 576/781 [00:36<00:12, 16.16it/s]
 74%|#######4  | 578/781 [00:36<00:12, 16.17it/s]
 74%|#######4  | 580/781 [00:36<00:12, 16.14it/s]
 75%|#######4  | 582/781 [00:36<00:12, 15.95it/s]
 75%|#######4  | 584/781 [00:36<00:12, 15.90it/s]
 75%|#######5  | 586/781 [00:36<00:12, 15.85it/s]
 75%|#######5  | 588/781 [00:37<00:12, 15.81it/s]
 76%|#######5  | 590/781 [00:37<00:12, 15.72it/s]
 76%|#######5  | 592/781 [00:37<00:11, 15.83it/s]
 76%|#######6  | 594/781 [00:37<00:11, 15.91it/s]
 76%|#######6  | 596/781 [00:37<00:11, 15.96it/s]
 77%|#######6  | 598/781 [00:37<00:11, 15.98it/s]
 77%|#######6  | 600/781 [00:37<00:11, 16.00it/s]
 77%|#######7  | 602/781 [00:37<00:11, 16.05it/s]
 77%|#######7  | 604/781 [00:38<00:10, 16.12it/s]
 78%|#######7  | 606/781 [00:38<00:10, 16.17it/s]
 78%|#######7  | 608/781 [00:38<00:10, 16.14it/s]
 78%|#######8  | 610/781 [00:38<00:10, 16.00it/s]
 78%|#######8  | 612/781 [00:38<00:10, 15.92it/s]
 79%|#######8  | 614/781 [00:38<00:10, 15.96it/s]
 79%|#######8  | 616/781 [00:38<00:10, 16.00it/s]
 79%|#######9  | 618/781 [00:38<00:10, 15.98it/s]
 79%|#######9  | 620/781 [00:39<00:10, 15.96it/s]
 80%|#######9  | 622/781 [00:39<00:09, 15.92it/s]
 80%|#######9  | 624/781 [00:39<00:09, 15.95it/s]
 80%|########  | 626/781 [00:39<00:09, 15.82it/s]
 80%|########  | 628/781 [00:39<00:09, 15.77it/s]
 81%|########  | 630/781 [00:39<00:09, 15.59it/s]
 81%|########  | 632/781 [00:39<00:09, 15.61it/s]
 81%|########1 | 634/781 [00:39<00:09, 15.60it/s]
 81%|########1 | 636/781 [00:40<00:09, 15.59it/s]
 82%|########1 | 638/781 [00:40<00:09, 15.64it/s]
 82%|########1 | 640/781 [00:40<00:08, 15.82it/s]
 82%|########2 | 642/781 [00:40<00:08, 15.96it/s]
 82%|########2 | 644/781 [00:40<00:08, 16.19it/s]
 83%|########2 | 646/781 [00:40<00:08, 16.08it/s]
 83%|########2 | 648/781 [00:40<00:08, 16.05it/s]
 83%|########3 | 650/781 [00:40<00:08, 16.04it/s]
 83%|########3 | 652/781 [00:41<00:08, 16.03it/s]
 84%|########3 | 654/781 [00:41<00:07, 16.06it/s]
 84%|########3 | 656/781 [00:41<00:07, 16.17it/s]
 84%|########4 | 658/781 [00:41<00:07, 16.37it/s]
 85%|########4 | 660/781 [00:41<00:07, 16.17it/s]
 85%|########4 | 662/781 [00:41<00:07, 16.14it/s]
 85%|########5 | 664/781 [00:41<00:07, 16.24it/s]
 85%|########5 | 666/781 [00:41<00:07, 16.33it/s]
 86%|########5 | 668/781 [00:42<00:06, 16.22it/s]
 86%|########5 | 670/781 [00:42<00:06, 16.18it/s]
 86%|########6 | 672/781 [00:42<00:06, 16.06it/s]
 86%|########6 | 674/781 [00:42<00:06, 16.08it/s]
 87%|########6 | 676/781 [00:42<00:06, 15.97it/s]
 87%|########6 | 678/781 [00:42<00:06, 16.01it/s]
 87%|########7 | 680/781 [00:42<00:06, 16.14it/s]
 87%|########7 | 682/781 [00:42<00:06, 16.11it/s]
 88%|########7 | 684/781 [00:43<00:06, 16.00it/s]
 88%|########7 | 686/781 [00:43<00:05, 16.08it/s]
 88%|########8 | 688/781 [00:43<00:05, 16.04it/s]
 88%|########8 | 690/781 [00:43<00:05, 15.96it/s]
 89%|########8 | 692/781 [00:43<00:05, 16.09it/s]
 89%|########8 | 694/781 [00:43<00:05, 16.07it/s]
 89%|########9 | 696/781 [00:43<00:05, 15.98it/s]
 89%|########9 | 698/781 [00:43<00:05, 16.03it/s]
 90%|########9 | 700/781 [00:44<00:05, 16.04it/s]
 90%|########9 | 702/781 [00:44<00:04, 16.15it/s]
 90%|######### | 704/781 [00:44<00:04, 16.01it/s]
 90%|######### | 706/781 [00:44<00:04, 15.91it/s]
 91%|######### | 708/781 [00:44<00:04, 15.87it/s]
 91%|######### | 710/781 [00:44<00:04, 15.84it/s]
 91%|#########1| 712/781 [00:44<00:04, 15.85it/s]
 91%|#########1| 714/781 [00:44<00:04, 15.80it/s]
 92%|#########1| 716/781 [00:45<00:04, 15.97it/s]
 92%|#########1| 718/781 [00:45<00:03, 15.95it/s]
 92%|#########2| 720/781 [00:45<00:03, 15.92it/s]
 92%|#########2| 722/781 [00:45<00:03, 15.92it/s]
 93%|#########2| 724/781 [00:45<00:03, 15.90it/s]
 93%|#########2| 726/781 [00:45<00:03, 15.85it/s]
 93%|#########3| 728/781 [00:45<00:04, 12.30it/s]
 93%|#########3| 730/781 [00:46<00:03, 13.22it/s]
 94%|#########3| 732/781 [00:46<00:03, 13.95it/s]
 94%|#########3| 734/781 [00:46<00:03, 14.63it/s]
 94%|#########4| 736/781 [00:46<00:02, 15.15it/s]
 94%|#########4| 738/781 [00:46<00:02, 15.41it/s]
 95%|#########4| 740/781 [00:46<00:02, 15.59it/s]
 95%|#########5| 742/781 [00:46<00:02, 15.72it/s]
 95%|#########5| 744/781 [00:46<00:02, 15.80it/s]
 96%|#########5| 746/781 [00:47<00:02, 15.87it/s]
 96%|#########5| 748/781 [00:47<00:02, 15.89it/s]
 96%|#########6| 750/781 [00:47<00:01, 15.95it/s]
 96%|#########6| 752/781 [00:47<00:01, 16.03it/s]
 97%|#########6| 754/781 [00:47<00:01, 16.13it/s]
 97%|#########6| 756/781 [00:47<00:01, 16.21it/s]
 97%|#########7| 758/781 [00:47<00:01, 16.19it/s]
 97%|#########7| 760/781 [00:47<00:01, 16.04it/s]
 98%|#########7| 762/781 [00:48<00:01, 15.93it/s]
 98%|#########7| 764/781 [00:48<00:01, 15.98it/s]
 98%|#########8| 766/781 [00:48<00:00, 15.98it/s]
 98%|#########8| 768/781 [00:48<00:00, 16.09it/s]
 99%|#########8| 770/781 [00:48<00:00, 16.12it/s]
 99%|#########8| 772/781 [00:48<00:00, 16.01it/s]
 99%|#########9| 774/781 [00:48<00:00, 16.04it/s]
 99%|#########9| 776/781 [00:48<00:00, 15.95it/s]
100%|#########9| 778/781 [00:49<00:00, 15.93it/s]
100%|#########9| 780/781 [00:49<00:00, 15.93it/s]
100%|##########| 781/781 [00:49<00:00, 15.87it/s]
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.26s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.26s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.26s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:54,  1.27s/it]
Validation Progress:   2%|2         | 1/44 [00:01<00:56,  1.32s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:49,  1.18s/it]
Validation Progress:   5%|4         | 2/44 [00:02<00:50,  1.20s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.25s/it]
Validation Progress:   7%|6         | 3/44 [00:03<00:51,  1.26s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.21s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:   9%|9         | 4/44 [00:04<00:48,  1.22s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:05<00:45,  1.17s/it]
Validation Progress:  11%|#1        | 5/44 [00:06<00:45,  1.17s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.16s/it]
Validation Progress:  14%|#3        | 6/44 [00:07<00:44,  1.17s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:42,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.16s/it]
Validation Progress:  16%|#5        | 7/44 [00:08<00:43,  1.19s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:42,  1.17s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:41,  1.16s/it]
Validation Progress:  18%|#8        | 8/44 [00:09<00:43,  1.21s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:40,  1.16s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:40,  1.16s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:40,  1.17s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:40,  1.17s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:40,  1.17s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:40,  1.17s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:41,  1.19s/it]
Validation Progress:  20%|##        | 9/44 [00:10<00:42,  1.21s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:39,  1.17s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:39,  1.18s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:40,  1.20s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.21s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.21s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:41,  1.22s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:40,  1.20s/it]
Validation Progress:  23%|##2       | 10/44 [00:11<00:40,  1.20s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.20s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.21s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.21s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.22s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:40,  1.21s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.20s/it]
Validation Progress:  25%|##5       | 11/44 [00:13<00:39,  1.20s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:38,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:38,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:38,  1.21s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.22s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:38,  1.21s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:38,  1.21s/it]
Validation Progress:  27%|##7       | 12/44 [00:14<00:39,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:37,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:37,  1.22s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.24s/it]
Validation Progress:  30%|##9       | 13/44 [00:15<00:38,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.22s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.22s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.23s/it]
Validation Progress:  32%|###1      | 14/44 [00:16<00:36,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.22s/it]
Validation Progress:  34%|###4      | 15/44 [00:18<00:35,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.23s/it]
Validation Progress:  36%|###6      | 16/44 [00:19<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:20<00:34,  1.27s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:21<00:32,  1.26s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.29s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:23<00:32,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:24<00:30,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:25<00:28,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:26<00:27,  1.23s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  52%|#####2    | 23/44 [00:28<00:26,  1.27s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  55%|#####4    | 24/44 [00:29<00:25,  1.29s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  57%|#####6    | 25/44 [00:30<00:25,  1.32s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [00:32<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.27s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [00:33<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:34<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [00:36<00:19,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:37<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:38<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [00:40<00:15,  1.33s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [00:41<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [00:42<00:12,  1.27s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [00:43<00:11,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [00:45<00:10,  1.30s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  84%|########4 | 37/44 [00:46<00:09,  1.34s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [00:47<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:49<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:50<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [00:51<00:04,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:53<00:02,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]
Validation Progress:  98%|#########7| 43/44 [00:54<00:01,  1.33s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.89s/it]
Validation Progress: 100%|##########| 44/44 [00:57<00:00,  1.31s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:02<00:00,  2.37s/it]
100%|##########| 1/1 [00:02<00:00,  2.37s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.61it/s]
  3%|2         | 4/157 [00:00<00:10, 14.62it/s]
  4%|3         | 6/157 [00:00<00:09, 15.20it/s]
  5%|5         | 8/157 [00:00<00:09, 15.36it/s]
  6%|6         | 10/157 [00:00<00:09, 15.46it/s]
  8%|7         | 12/157 [00:00<00:09, 15.51it/s]
  9%|8         | 14/157 [00:00<00:09, 15.61it/s]
 10%|#         | 16/157 [00:01<00:09, 15.60it/s]
 11%|#1        | 18/157 [00:01<00:08, 15.68it/s]
 13%|#2        | 20/157 [00:01<00:08, 15.78it/s]
 14%|#4        | 22/157 [00:01<00:08, 15.86it/s]
 15%|#5        | 24/157 [00:01<00:08, 15.84it/s]
 17%|#6        | 26/157 [00:01<00:08, 15.80it/s]
 18%|#7        | 28/157 [00:01<00:08, 15.75it/s]
 19%|#9        | 30/157 [00:01<00:08, 15.70it/s]
 20%|##        | 32/157 [00:02<00:07, 15.72it/s]
 22%|##1       | 34/157 [00:02<00:07, 15.70it/s]
 23%|##2       | 36/157 [00:02<00:07, 15.64it/s]
 24%|##4       | 38/157 [00:02<00:07, 15.55it/s]
 25%|##5       | 40/157 [00:02<00:07, 15.48it/s]
 27%|##6       | 42/157 [00:02<00:07, 15.51it/s]
 28%|##8       | 44/157 [00:02<00:07, 15.47it/s]
 29%|##9       | 46/157 [00:02<00:07, 15.35it/s]
 31%|###       | 48/157 [00:03<00:07, 15.40it/s]
 32%|###1      | 50/157 [00:03<00:06, 15.39it/s]
 33%|###3      | 52/157 [00:03<00:06, 15.34it/s]
 34%|###4      | 54/157 [00:03<00:06, 15.31it/s]
 36%|###5      | 56/157 [00:03<00:06, 15.27it/s]
 37%|###6      | 58/157 [00:03<00:06, 15.36it/s]
 38%|###8      | 60/157 [00:03<00:06, 15.38it/s]
 39%|###9      | 62/157 [00:04<00:06, 15.22it/s]
 41%|####      | 64/157 [00:04<00:06, 15.35it/s]
 42%|####2     | 66/157 [00:04<00:05, 15.33it/s]
 43%|####3     | 68/157 [00:04<00:05, 15.46it/s]
 45%|####4     | 70/157 [00:04<00:05, 15.56it/s]
 46%|####5     | 72/157 [00:04<00:05, 15.58it/s]
 47%|####7     | 74/157 [00:04<00:05, 15.52it/s]
 48%|####8     | 76/157 [00:04<00:05, 15.54it/s]
 50%|####9     | 78/157 [00:05<00:05, 15.57it/s]
 51%|#####     | 80/157 [00:05<00:04, 15.55it/s]
 52%|#####2    | 82/157 [00:05<00:04, 15.55it/s]
 54%|#####3    | 84/157 [00:05<00:04, 15.56it/s]
 55%|#####4    | 86/157 [00:05<00:04, 15.63it/s]
 56%|#####6    | 88/157 [00:05<00:04, 15.65it/s]
 57%|#####7    | 90/157 [00:05<00:04, 15.66it/s]
 59%|#####8    | 92/157 [00:05<00:04, 15.64it/s]
 60%|#####9    | 94/157 [00:06<00:04, 15.67it/s]
 61%|######1   | 96/157 [00:06<00:03, 15.64it/s]
 62%|######2   | 98/157 [00:06<00:03, 15.55it/s]
 64%|######3   | 100/157 [00:06<00:03, 15.45it/s]
 65%|######4   | 102/157 [00:06<00:03, 15.35it/s]
 66%|######6   | 104/157 [00:06<00:03, 15.38it/s]
 68%|######7   | 106/157 [00:06<00:03, 15.44it/s]
 69%|######8   | 108/157 [00:06<00:03, 15.48it/s]
 70%|#######   | 110/157 [00:07<00:03, 15.40it/s]
 71%|#######1  | 112/157 [00:07<00:02, 15.42it/s]
 73%|#######2  | 114/157 [00:07<00:02, 15.45it/s]
 74%|#######3  | 116/157 [00:07<00:02, 15.40it/s]
 75%|#######5  | 118/157 [00:07<00:02, 15.45it/s]
 76%|#######6  | 120/157 [00:07<00:02, 15.37it/s]
 78%|#######7  | 122/157 [00:07<00:02, 15.39it/s]
 79%|#######8  | 124/157 [00:08<00:02, 15.29it/s]
 80%|########  | 126/157 [00:08<00:02, 15.36it/s]
 82%|########1 | 128/157 [00:08<00:01, 15.89it/s]
 83%|########2 | 130/157 [00:08<00:01, 16.22it/s]
 84%|########4 | 132/157 [00:08<00:01, 16.27it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.77it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.66it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.40it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.47it/s]
 90%|######### | 142/157 [00:09<00:00, 16.14it/s]
 92%|#########1| 144/157 [00:09<00:00, 16.12it/s]
 93%|#########2| 146/157 [00:09<00:00, 16.61it/s]
 94%|#########4| 148/157 [00:09<00:00, 16.38it/s]
 96%|#########5| 150/157 [00:09<00:00, 16.13it/s]
 97%|#########6| 152/157 [00:09<00:00, 15.93it/s]
 98%|#########8| 154/157 [00:09<00:00, 15.84it/s]
 99%|#########9| 156/157 [00:09<00:00, 15.97it/s]
100%|##########| 157/157 [00:09<00:00, 15.70it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:43, 17.72it/s]
  1%|          | 4/781 [00:00<00:41, 18.91it/s]
  1%|          | 6/781 [00:00<00:42, 18.32it/s]
  1%|1         | 8/781 [00:00<00:40, 18.89it/s]
  1%|1         | 11/781 [00:00<00:39, 19.27it/s]
  2%|1         | 13/781 [00:00<00:41, 18.68it/s]
  2%|1         | 15/781 [00:00<00:41, 18.60it/s]
  2%|2         | 18/781 [00:00<00:40, 19.05it/s]
  3%|2         | 21/781 [00:01<00:38, 19.51it/s]
  3%|3         | 24/781 [00:01<00:38, 19.80it/s]
  3%|3         | 26/781 [00:01<00:39, 19.27it/s]
  4%|3         | 28/781 [00:01<00:39, 18.98it/s]
  4%|3         | 31/781 [00:01<00:38, 19.44it/s]
  4%|4         | 33/781 [00:01<00:38, 19.38it/s]
  4%|4         | 35/781 [00:01<00:39, 18.81it/s]
  5%|4         | 37/781 [00:01<00:39, 18.76it/s]
  5%|4         | 39/781 [00:02<00:38, 19.04it/s]
  5%|5         | 41/781 [00:02<00:39, 18.62it/s]
  6%|5         | 43/781 [00:02<00:40, 18.31it/s]
  6%|5         | 45/781 [00:02<00:40, 18.04it/s]
  6%|6         | 47/781 [00:02<00:41, 17.74it/s]
  6%|6         | 49/781 [00:02<00:41, 17.54it/s]
  7%|6         | 51/781 [00:02<00:41, 17.53it/s]
  7%|6         | 53/781 [00:02<00:41, 17.46it/s]
  7%|7         | 55/781 [00:02<00:41, 17.39it/s]
  7%|7         | 57/781 [00:03<00:41, 17.39it/s]
  8%|7         | 59/781 [00:03<00:41, 17.41it/s]
  8%|7         | 61/781 [00:03<00:41, 17.41it/s]
  8%|8         | 63/781 [00:03<00:41, 17.40it/s]
  8%|8         | 65/781 [00:03<00:41, 17.30it/s]
  9%|8         | 67/781 [00:03<00:41, 17.33it/s]
  9%|8         | 69/781 [00:03<00:40, 17.41it/s]
  9%|9         | 71/781 [00:03<00:40, 17.39it/s]
  9%|9         | 73/781 [00:04<00:40, 17.39it/s]
 10%|9         | 75/781 [00:04<00:40, 17.38it/s]
 10%|9         | 77/781 [00:04<00:40, 17.37it/s]
 10%|#         | 79/781 [00:04<00:40, 17.35it/s]
 10%|#         | 81/781 [00:04<00:40, 17.37it/s]
 11%|#         | 83/781 [00:04<00:40, 17.40it/s]
 11%|#         | 85/781 [00:04<00:39, 17.42it/s]
 11%|#1        | 87/781 [00:04<00:39, 17.46it/s]
 11%|#1        | 89/781 [00:04<00:39, 17.41it/s]
 12%|#1        | 91/781 [00:05<00:39, 17.44it/s]
 12%|#1        | 93/781 [00:05<00:39, 17.46it/s]
 12%|#2        | 95/781 [00:05<00:39, 17.45it/s]
 12%|#2        | 97/781 [00:05<00:39, 17.41it/s]
 13%|#2        | 99/781 [00:05<00:39, 17.40it/s]
 13%|#2        | 101/781 [00:05<00:39, 17.39it/s]
 13%|#3        | 103/781 [00:05<00:39, 17.36it/s]
 13%|#3        | 105/781 [00:05<00:38, 17.35it/s]
 14%|#3        | 107/781 [00:05<00:38, 17.38it/s]
 14%|#3        | 109/781 [00:06<00:38, 17.41it/s]
 14%|#4        | 111/781 [00:06<00:38, 17.45it/s]
 14%|#4        | 113/781 [00:06<00:38, 17.25it/s]
 15%|#4        | 115/781 [00:06<00:38, 17.19it/s]
 15%|#4        | 117/781 [00:06<00:38, 17.07it/s]
 15%|#5        | 119/781 [00:06<00:37, 17.60it/s]
 16%|#5        | 122/781 [00:06<00:35, 18.52it/s]
 16%|#5        | 124/781 [00:06<00:35, 18.58it/s]
 16%|#6        | 126/781 [00:07<00:35, 18.53it/s]
 16%|#6        | 128/781 [00:07<00:35, 18.43it/s]
 17%|#6        | 131/781 [00:07<00:33, 19.23it/s]
 17%|#7        | 133/781 [00:07<00:33, 19.18it/s]
 17%|#7        | 135/781 [00:07<00:34, 18.96it/s]
 18%|#7        | 137/781 [00:07<00:33, 19.02it/s]
 18%|#7        | 139/781 [00:07<00:34, 18.69it/s]
 18%|#8        | 142/781 [00:07<00:33, 19.05it/s]
 18%|#8        | 144/781 [00:07<00:33, 18.74it/s]
 19%|#8        | 146/781 [00:08<00:33, 18.89it/s]
 19%|#8        | 148/781 [00:08<00:33, 19.15it/s]
 19%|#9        | 151/781 [00:08<00:32, 19.59it/s]
 20%|#9        | 153/781 [00:08<00:32, 19.37it/s]
 20%|#9        | 156/781 [00:08<00:31, 19.75it/s]
 20%|##        | 158/781 [00:08<00:31, 19.51it/s]
 20%|##        | 160/781 [00:08<00:32, 19.21it/s]
 21%|##        | 162/781 [00:08<00:32, 18.94it/s]
 21%|##1       | 165/781 [00:09<00:31, 19.49it/s]
 21%|##1       | 167/781 [00:09<00:31, 19.32it/s]
 22%|##1       | 170/781 [00:09<00:31, 19.69it/s]
 22%|##2       | 172/781 [00:09<00:31, 19.59it/s]
 22%|##2       | 174/781 [00:09<00:31, 19.22it/s]
 23%|##2       | 177/781 [00:09<00:30, 19.64it/s]
 23%|##2       | 179/781 [00:09<00:30, 19.50it/s]
 23%|##3       | 181/781 [00:09<00:30, 19.61it/s]
 24%|##3       | 184/781 [00:10<00:30, 19.69it/s]
 24%|##3       | 187/781 [00:10<00:29, 19.92it/s]
 24%|##4       | 190/781 [00:10<00:29, 20.20it/s]
 25%|##4       | 193/781 [00:10<00:29, 20.05it/s]
 25%|##5       | 196/781 [00:10<00:29, 20.05it/s]
 25%|##5       | 199/781 [00:10<00:29, 19.96it/s]
 26%|##5       | 201/781 [00:10<00:29, 19.43it/s]
 26%|##5       | 203/781 [00:10<00:30, 19.00it/s]
 26%|##6       | 205/781 [00:11<00:30, 18.68it/s]
 27%|##6       | 207/781 [00:11<00:30, 18.79it/s]
 27%|##6       | 209/781 [00:11<00:30, 18.69it/s]
 27%|##7       | 211/781 [00:11<00:30, 18.40it/s]
 27%|##7       | 213/781 [00:11<00:30, 18.49it/s]
 28%|##7       | 215/781 [00:11<00:30, 18.61it/s]
 28%|##7       | 217/781 [00:11<00:30, 18.36it/s]
 28%|##8       | 219/781 [00:11<00:30, 18.18it/s]
 28%|##8       | 221/781 [00:11<00:31, 18.04it/s]
 29%|##8       | 223/781 [00:12<00:31, 17.95it/s]
 29%|##8       | 225/781 [00:12<00:30, 18.31it/s]
 29%|##9       | 227/781 [00:12<00:30, 18.14it/s]
 29%|##9       | 229/781 [00:12<00:30, 18.01it/s]
 30%|##9       | 231/781 [00:12<00:30, 18.33it/s]
 30%|##9       | 233/781 [00:12<00:30, 18.18it/s]
 30%|###       | 235/781 [00:12<00:30, 18.07it/s]
 30%|###       | 237/781 [00:12<00:29, 18.40it/s]
 31%|###       | 240/781 [00:12<00:28, 19.15it/s]
 31%|###       | 242/781 [00:13<00:28, 19.21it/s]
 31%|###1      | 245/781 [00:13<00:27, 19.79it/s]
 32%|###1      | 247/781 [00:13<00:27, 19.77it/s]
 32%|###1      | 249/781 [00:13<00:27, 19.61it/s]
 32%|###2      | 252/781 [00:13<00:26, 19.91it/s]
 33%|###2      | 255/781 [00:13<00:25, 20.27it/s]
 33%|###3      | 258/781 [00:13<00:26, 20.07it/s]
 33%|###3      | 261/781 [00:14<00:26, 19.96it/s]
 34%|###3      | 263/781 [00:14<00:26, 19.91it/s]
 34%|###3      | 265/781 [00:14<00:26, 19.79it/s]
 34%|###4      | 268/781 [00:14<00:25, 20.01it/s]
 35%|###4      | 270/781 [00:14<00:25, 19.79it/s]
 35%|###4      | 273/781 [00:14<00:25, 20.09it/s]
 35%|###5      | 276/781 [00:14<00:24, 20.36it/s]
 36%|###5      | 279/781 [00:14<00:24, 20.14it/s]
 36%|###6      | 282/781 [00:15<00:24, 20.11it/s]
 36%|###6      | 285/781 [00:15<00:24, 20.22it/s]
 37%|###6      | 288/781 [00:15<00:24, 20.20it/s]
 37%|###7      | 291/781 [00:15<00:24, 20.30it/s]
 38%|###7      | 294/781 [00:15<00:24, 19.78it/s]
 38%|###8      | 297/781 [00:15<00:24, 19.88it/s]
 38%|###8      | 300/781 [00:15<00:23, 20.07it/s]
 39%|###8      | 303/781 [00:16<00:23, 19.92it/s]
 39%|###9      | 306/781 [00:16<00:23, 20.20it/s]
 40%|###9      | 309/781 [00:16<00:23, 20.22it/s]
 40%|###9      | 312/781 [00:16<00:23, 20.07it/s]
 40%|####      | 315/781 [00:16<00:23, 19.98it/s]
 41%|####      | 318/781 [00:16<00:23, 20.12it/s]
 41%|####1     | 321/781 [00:17<00:22, 20.24it/s]
 41%|####1     | 324/781 [00:17<00:22, 20.08it/s]
 42%|####1     | 327/781 [00:17<00:22, 19.85it/s]
 42%|####2     | 329/781 [00:17<00:23, 19.63it/s]
 42%|####2     | 331/781 [00:17<00:23, 19.23it/s]
 43%|####2     | 333/781 [00:17<00:23, 19.39it/s]
 43%|####2     | 335/781 [00:17<00:23, 18.87it/s]
 43%|####3     | 337/781 [00:17<00:23, 18.91it/s]
 43%|####3     | 339/781 [00:17<00:23, 19.19it/s]
 44%|####3     | 341/781 [00:18<00:23, 18.99it/s]
 44%|####4     | 344/781 [00:18<00:22, 19.41it/s]
 44%|####4     | 346/781 [00:18<00:22, 19.49it/s]
 45%|####4     | 349/781 [00:18<00:21, 19.77it/s]
 45%|####4     | 351/781 [00:18<00:22, 19.33it/s]
 45%|####5     | 354/781 [00:18<00:21, 19.71it/s]
 46%|####5     | 356/781 [00:18<00:21, 19.71it/s]
 46%|####5     | 358/781 [00:18<00:21, 19.45it/s]
 46%|####6     | 360/781 [00:19<00:21, 19.16it/s]
 46%|####6     | 362/781 [00:19<00:21, 19.33it/s]
 47%|####6     | 365/781 [00:19<00:20, 19.90it/s]
 47%|####7     | 368/781 [00:19<00:20, 20.31it/s]
 48%|####7     | 371/781 [00:19<00:19, 20.59it/s]
 48%|####7     | 374/781 [00:19<00:19, 20.50it/s]
 48%|####8     | 377/781 [00:19<00:20, 20.15it/s]
 49%|####8     | 380/781 [00:20<00:19, 20.05it/s]
 49%|####9     | 383/781 [00:20<00:19, 20.32it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.61it/s]
 50%|####9     | 388/781 [00:20<00:20, 19.14it/s]
 50%|#####     | 391/781 [00:20<00:19, 19.56it/s]
 50%|#####     | 393/781 [00:20<00:19, 19.50it/s]
 51%|#####     | 396/781 [00:20<00:19, 19.77it/s]
 51%|#####1    | 399/781 [00:20<00:19, 19.98it/s]
 51%|#####1    | 401/781 [00:21<00:19, 19.82it/s]
 52%|#####1    | 403/781 [00:21<00:19, 19.16it/s]
 52%|#####1    | 405/781 [00:21<00:19, 19.02it/s]
 52%|#####2    | 407/781 [00:21<00:19, 18.81it/s]
 52%|#####2    | 409/781 [00:21<00:19, 18.79it/s]
 53%|#####2    | 412/781 [00:21<00:18, 19.48it/s]
 53%|#####3    | 415/781 [00:21<00:18, 19.99it/s]
 54%|#####3    | 418/781 [00:21<00:17, 20.31it/s]
 54%|#####3    | 421/781 [00:22<00:17, 20.12it/s]
 54%|#####4    | 424/781 [00:22<00:17, 20.15it/s]
 55%|#####4    | 427/781 [00:22<00:18, 19.51it/s]
 55%|#####5    | 430/781 [00:22<00:17, 19.78it/s]
 55%|#####5    | 432/781 [00:22<00:17, 19.78it/s]
 56%|#####5    | 434/781 [00:22<00:17, 19.57it/s]
 56%|#####5    | 436/781 [00:22<00:18, 18.96it/s]
 56%|#####6    | 439/781 [00:23<00:17, 19.45it/s]
 57%|#####6    | 442/781 [00:23<00:17, 19.94it/s]
 57%|#####6    | 445/781 [00:23<00:16, 20.26it/s]
 57%|#####7    | 448/781 [00:23<00:16, 20.11it/s]
 58%|#####7    | 451/781 [00:23<00:16, 20.45it/s]
 58%|#####8    | 454/781 [00:23<00:15, 20.64it/s]
 59%|#####8    | 457/781 [00:23<00:15, 20.76it/s]
 59%|#####8    | 460/781 [00:24<00:15, 20.63it/s]
 59%|#####9    | 463/781 [00:24<00:15, 20.60it/s]
 60%|#####9    | 466/781 [00:24<00:15, 20.49it/s]
 60%|######    | 469/781 [00:24<00:15, 20.16it/s]
 60%|######    | 472/781 [00:24<00:15, 20.39it/s]
 61%|######    | 475/781 [00:24<00:14, 20.57it/s]
 61%|######1   | 478/781 [00:24<00:14, 20.35it/s]
 62%|######1   | 481/781 [00:25<00:14, 20.16it/s]
 62%|######1   | 484/781 [00:25<00:14, 20.39it/s]
 62%|######2   | 487/781 [00:25<00:14, 20.57it/s]
 63%|######2   | 490/781 [00:25<00:14, 20.69it/s]
 63%|######3   | 493/781 [00:25<00:13, 20.81it/s]
 64%|######3   | 496/781 [00:25<00:13, 20.88it/s]
 64%|######3   | 499/781 [00:25<00:13, 20.96it/s]
 64%|######4   | 502/781 [00:26<00:13, 20.98it/s]
 65%|######4   | 505/781 [00:26<00:13, 20.59it/s]
 65%|######5   | 508/781 [00:26<00:13, 20.45it/s]
 65%|######5   | 511/781 [00:26<00:13, 20.46it/s]
 66%|######5   | 514/781 [00:26<00:12, 20.58it/s]
 66%|######6   | 517/781 [00:26<00:12, 20.31it/s]
 67%|######6   | 520/781 [00:26<00:12, 20.17it/s]
 67%|######6   | 523/781 [00:27<00:12, 19.86it/s]
 67%|######7   | 525/781 [00:27<00:12, 19.72it/s]
 67%|######7   | 527/781 [00:27<00:13, 19.20it/s]
 68%|######7   | 529/781 [00:27<00:13, 18.76it/s]
 68%|######7   | 531/781 [00:27<00:13, 18.71it/s]
 68%|######8   | 534/781 [00:27<00:12, 19.28it/s]
 69%|######8   | 536/781 [00:27<00:12, 19.36it/s]
 69%|######8   | 538/781 [00:27<00:12, 19.28it/s]
 69%|######9   | 541/781 [00:28<00:12, 19.65it/s]
 70%|######9   | 544/781 [00:28<00:11, 20.05it/s]
 70%|#######   | 547/781 [00:28<00:11, 20.08it/s]
 70%|#######   | 550/781 [00:28<00:11, 20.21it/s]
 71%|#######   | 553/781 [00:28<00:11, 20.06it/s]
 71%|#######1  | 556/781 [00:28<00:11, 20.28it/s]
 72%|#######1  | 559/781 [00:28<00:10, 20.22it/s]
 72%|#######1  | 562/781 [00:29<00:10, 20.30it/s]
 72%|#######2  | 565/781 [00:29<00:10, 20.49it/s]
 73%|#######2  | 568/781 [00:29<00:10, 20.39it/s]
 73%|#######3  | 571/781 [00:29<00:10, 20.10it/s]
 73%|#######3  | 574/781 [00:29<00:10, 20.34it/s]
 74%|#######3  | 577/781 [00:29<00:09, 20.52it/s]
 74%|#######4  | 580/781 [00:29<00:09, 20.65it/s]
 75%|#######4  | 583/781 [00:30<00:09, 20.17it/s]
 75%|#######5  | 586/781 [00:30<00:09, 19.92it/s]
 75%|#######5  | 589/781 [00:30<00:09, 20.21it/s]
 76%|#######5  | 592/781 [00:30<00:09, 20.43it/s]
 76%|#######6  | 595/781 [00:30<00:09, 19.89it/s]
 77%|#######6  | 598/781 [00:30<00:09, 20.17it/s]
 77%|#######6  | 601/781 [00:31<00:08, 20.18it/s]
 77%|#######7  | 604/781 [00:31<00:08, 20.32it/s]
 78%|#######7  | 607/781 [00:31<00:08, 20.07it/s]
 78%|#######8  | 610/781 [00:31<00:08, 20.32it/s]
 78%|#######8  | 613/781 [00:31<00:08, 20.51it/s]
 79%|#######8  | 616/781 [00:31<00:08, 20.62it/s]
 79%|#######9  | 619/781 [00:31<00:07, 20.26it/s]
 80%|#######9  | 622/781 [00:32<00:07, 20.49it/s]
 80%|########  | 625/781 [00:32<00:07, 20.07it/s]
 80%|########  | 628/781 [00:32<00:07, 19.93it/s]
 81%|########  | 630/781 [00:32<00:07, 19.82it/s]
 81%|########  | 632/781 [00:32<00:07, 19.67it/s]
 81%|########1 | 635/781 [00:32<00:07, 20.09it/s]
 82%|########1 | 638/781 [00:32<00:07, 19.64it/s]
 82%|########2 | 641/781 [00:33<00:07, 19.80it/s]
 82%|########2 | 644/781 [00:33<00:06, 20.02it/s]
 83%|########2 | 647/781 [00:33<00:06, 19.94it/s]
 83%|########3 | 649/781 [00:33<00:06, 19.76it/s]
 83%|########3 | 652/781 [00:33<00:06, 19.85it/s]
 84%|########3 | 654/781 [00:33<00:06, 19.50it/s]
 84%|########4 | 657/781 [00:33<00:06, 19.76it/s]
 84%|########4 | 659/781 [00:33<00:06, 19.62it/s]
 85%|########4 | 661/781 [00:34<00:06, 19.63it/s]
 85%|########4 | 663/781 [00:34<00:06, 19.49it/s]
 85%|########5 | 665/781 [00:34<00:05, 19.56it/s]
 86%|########5 | 668/781 [00:34<00:05, 20.04it/s]
 86%|########5 | 670/781 [00:34<00:05, 19.99it/s]
 86%|########6 | 672/781 [00:34<00:05, 19.78it/s]
 86%|########6 | 675/781 [00:34<00:05, 20.03it/s]
 87%|########6 | 677/781 [00:34<00:05, 19.97it/s]
 87%|########6 | 679/781 [00:34<00:05, 19.59it/s]
 87%|########7 | 681/781 [00:35<00:05, 17.75it/s]
 88%|########7 | 684/781 [00:35<00:05, 18.82it/s]
 88%|########7 | 687/781 [00:35<00:04, 19.54it/s]
 88%|########8 | 690/781 [00:35<00:04, 20.03it/s]
 89%|########8 | 693/781 [00:35<00:04, 20.33it/s]
 89%|########9 | 696/781 [00:35<00:04, 20.54it/s]
 90%|########9 | 699/781 [00:35<00:03, 20.54it/s]
 90%|########9 | 702/781 [00:36<00:03, 20.19it/s]
 90%|######### | 705/781 [00:36<00:03, 20.03it/s]
 91%|######### | 708/781 [00:36<00:03, 19.92it/s]
 91%|######### | 710/781 [00:36<00:03, 19.36it/s]
 91%|#########1| 712/781 [00:36<00:03, 19.42it/s]
 91%|#########1| 714/781 [00:36<00:03, 19.26it/s]
 92%|#########1| 716/781 [00:36<00:03, 19.22it/s]
 92%|#########1| 718/781 [00:36<00:03, 19.30it/s]
 92%|#########2| 720/781 [00:37<00:03, 18.77it/s]
 92%|#########2| 722/781 [00:37<00:03, 18.88it/s]
 93%|#########2| 724/781 [00:37<00:03, 18.99it/s]
 93%|#########2| 726/781 [00:37<00:02, 18.84it/s]
 93%|#########3| 728/781 [00:37<00:02, 19.10it/s]
 93%|#########3| 730/781 [00:37<00:02, 18.99it/s]
 94%|#########3| 732/781 [00:37<00:02, 18.73it/s]
 94%|#########3| 734/781 [00:37<00:02, 19.03it/s]
 94%|#########4| 737/781 [00:37<00:02, 19.74it/s]
 95%|#########4| 739/781 [00:38<00:02, 19.75it/s]
 95%|#########4| 741/781 [00:38<00:02, 19.63it/s]
 95%|#########5| 744/781 [00:38<00:01, 19.70it/s]
 96%|#########5| 747/781 [00:38<00:01, 19.79it/s]
 96%|#########5| 749/781 [00:38<00:01, 19.68it/s]
 96%|#########6| 752/781 [00:38<00:01, 19.75it/s]
 97%|#########6| 755/781 [00:38<00:01, 20.06it/s]
 97%|#########7| 758/781 [00:39<00:01, 19.42it/s]
 97%|#########7| 760/781 [00:39<00:01, 19.38it/s]
 98%|#########7| 762/781 [00:39<00:01, 18.98it/s]
 98%|#########7| 764/781 [00:39<00:00, 18.67it/s]
 98%|#########8| 766/781 [00:39<00:00, 18.81it/s]
 98%|#########8| 769/781 [00:39<00:00, 19.42it/s]
 99%|#########8| 771/781 [00:39<00:00, 19.53it/s]
 99%|#########8| 773/781 [00:39<00:00, 19.05it/s]
 99%|#########9| 775/781 [00:39<00:00, 18.96it/s]
 99%|#########9| 777/781 [00:40<00:00, 19.20it/s]
100%|#########9| 779/781 [00:40<00:00, 19.05it/s]
100%|##########| 781/781 [00:40<00:00, 19.42it/s]






=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================







=======================================================   RESTART [11-26 03:00:17]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241125_110343-lbrdfnai
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run spring-pyramid-549
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/lbrdfnai
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
AUTOTUNE mm(3420x28, 28x1920)
  triton_mm_9 0.0243 ms 100.0%
  triton_mm_4 0.0253 ms 96.0%
  triton_mm_7 0.0254 ms 95.7%
  triton_mm_11 0.0257 ms 94.6%
  triton_mm_6 0.0257 ms 94.6%
  triton_mm_12 0.0258 ms 94.4%
  triton_mm_8 0.0258 ms 94.4%
  triton_mm_3 0.0261 ms 93.3%
  triton_mm_16 0.0273 ms 89.2%
  triton_mm_2 0.0273 ms 89.0%
SingleProcess AUTOTUNE benchmarking takes 3.1705 seconds and 0.0000 seconds precompiling
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3724 ms 100.0%
  triton_mm_11504 0.4128 ms 90.2%
  triton_mm_11508 0.4233 ms 88.0%
  mm 0.4309 ms 86.4%
  triton_mm_11505 0.4483 ms 83.1%
  triton_mm_11506 0.4611 ms 80.8%
  triton_mm_11512 0.4629 ms 80.4%
  triton_mm_11509 0.4668 ms 79.8%
  triton_mm_11501 0.5605 ms 66.4%
  triton_mm_11502 0.5737 ms 64.9%
SingleProcess AUTOTUNE benchmarking takes 2.1536 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3929 ms 100.0%
  mm 0.4077 ms 96.4%
  triton_mm_11523 0.4387 ms 89.6%
  triton_mm_11527 0.4497 ms 87.4%
  triton_mm_11524 0.4713 ms 83.4%
  triton_mm_11525 0.4804 ms 81.8%
  triton_mm_11531 0.4889 ms 80.4%
  triton_mm_11528 0.4909 ms 80.0%
  triton_mm_11520 0.5897 ms 66.6%
  triton_mm_11521 0.6127 ms 64.1%
SingleProcess AUTOTUNE benchmarking takes 2.1679 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  bias_addmm 0.1760 ms 100.0%
  triton_mm_11549 0.1763 ms 99.8%
  triton_mm_11546 0.1788 ms 98.5%
  triton_mm_11542 0.1791 ms 98.2%
  addmm 0.1817 ms 96.9%
  triton_mm_11543 0.1923 ms 91.5%
  triton_mm_11547 0.1923 ms 91.5%
  triton_mm_11544 0.1965 ms 89.6%
  triton_mm_11550 0.2144 ms 82.1%
  triton_mm_11540 0.2350 ms 74.9%
SingleProcess AUTOTUNE benchmarking takes 2.2801 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:22, 119.12s/it]AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8678 0.0315 ms 100.0%
  triton_mm_8680 0.0315 ms 100.0%
  triton_mm_8676 0.0316 ms 99.7%
  triton_mm_8681 0.0319 ms 98.7%
  triton_mm_8675 0.0328 ms 96.1%
  triton_mm_8677 0.0328 ms 96.1%
  triton_mm_8673 0.0335 ms 94.2%
  triton_mm_8672 0.0349 ms 90.3%
  triton_mm_8684 0.0361 ms 87.3%
  triton_mm_8671 0.0362 ms 87.0%
SingleProcess AUTOTUNE benchmarking takes 2.2085 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1751 ms 100.0%
  bias_addmm 0.1764 ms 99.3%
  triton_mm_11542 0.1785 ms 98.1%
  triton_mm_11546 0.1793 ms 97.7%
  addmm 0.1822 ms 96.1%
  triton_mm_11543 0.1915 ms 91.4%
  triton_mm_11547 0.1929 ms 90.8%
  triton_mm_11544 0.1942 ms 90.2%
  triton_mm_11550 0.2137 ms 81.9%
  triton_mm_11540 0.2347 ms 74.6%
SingleProcess AUTOTUNE benchmarking takes 2.2124 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:22, 119.12s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:22, 119.12s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2689 ms 100.0%
  mm 1.4434 ms 87.9%
  triton_mm_8754 1.5425 ms 82.3%
  triton_mm_8758 1.5517 ms 81.8%
  triton_mm_8750 1.5521 ms 81.8%
  triton_mm_8755 1.6677 ms 76.1%
  triton_mm_8751 1.6681 ms 76.1%
  triton_mm_8752 1.6872 ms 75.2%
  triton_mm_8747 2.0695 ms 61.3%
  triton_mm_8748 2.1477 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.3208 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8943 1.2676 ms 100.0%
  mm 1.3013 ms 97.4%
  triton_mm_8940 1.5572 ms 81.4%
  triton_mm_8944 1.5675 ms 80.9%
  triton_mm_8936 1.5939 ms 79.5%
  triton_mm_8941 1.6812 ms 75.4%
  triton_mm_8937 1.7020 ms 74.5%
  triton_mm_8938 1.7029 ms 74.4%
  triton_mm_8933 2.0863 ms 60.8%
  triton_mm_8934 2.1431 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.2297 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11175 1.2692 ms 100.0%
  mm 1.2931 ms 98.2%
  triton_mm_11172 1.5604 ms 81.3%
  triton_mm_11176 1.5863 ms 80.0%
  triton_mm_11168 1.6148 ms 78.6%
  triton_mm_11173 1.6835 ms 75.4%
  triton_mm_11170 1.7107 ms 74.2%
  triton_mm_11169 1.7209 ms 73.8%
  triton_mm_11165 2.0861 ms 60.8%
  triton_mm_11166 2.1670 ms 58.6%
SingleProcess AUTOTUNE benchmarking takes 2.2351 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_11454 1.2688 ms 100.0%
  mm 1.3046 ms 97.3%
  triton_mm_11451 1.5574 ms 81.5%
  triton_mm_11455 1.5678 ms 80.9%
  triton_mm_11447 1.6327 ms 77.7%
  triton_mm_11452 1.6820 ms 75.4%
  triton_mm_11448 1.7014 ms 74.6%
  triton_mm_11449 1.7020 ms 74.5%
  triton_mm_11444 2.0873 ms 60.8%
  triton_mm_11445 2.1444 ms 59.2%
SingleProcess AUTOTUNE benchmarking takes 2.2344 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3752 ms 100.0%
  mm 0.3855 ms 97.3%
  triton_mm_11504 0.4179 ms 89.8%
  triton_mm_11508 0.4273 ms 87.8%
  triton_mm_11505 0.4494 ms 83.5%
  triton_mm_11506 0.4610 ms 81.4%
  triton_mm_11512 0.4637 ms 80.9%
  triton_mm_11509 0.4679 ms 80.2%
  triton_mm_11501 0.5640 ms 66.5%
  triton_mm_11502 0.5814 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.1502 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3966 ms 100.0%
  mm 0.4092 ms 96.9%
  triton_mm_11523 0.4385 ms 90.4%
  triton_mm_11527 0.4515 ms 87.8%
  triton_mm_11524 0.4707 ms 84.2%
  triton_mm_11525 0.4812 ms 82.4%
  triton_mm_11531 0.4911 ms 80.8%
  triton_mm_11528 0.4948 ms 80.2%
  triton_mm_11520 0.5929 ms 66.9%
  triton_mm_11521 0.6143 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.1535 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8677 0.0317 ms 100.0%
  triton_mm_8676 0.0318 ms 99.8%
  triton_mm_8680 0.0318 ms 99.8%
  triton_mm_8675 0.0331 ms 95.8%
  triton_mm_8678 0.0332 ms 95.8%
  triton_mm_8681 0.0335 ms 94.7%
  triton_mm_8673 0.0349 ms 90.8%
  triton_mm_8672 0.0351 ms 90.4%
  triton_mm_8684 0.0364 ms 87.1%
  triton_mm_8671 0.0365 ms 86.9%
SingleProcess AUTOTUNE benchmarking takes 2.2030 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:22, 119.12s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2714 ms 100.0%
  mm 1.4431 ms 88.1%
  triton_mm_8754 1.5417 ms 82.5%
  triton_mm_8750 1.5543 ms 81.8%
  triton_mm_8758 1.5577 ms 81.6%
  triton_mm_8755 1.6814 ms 75.6%
  triton_mm_8751 1.6820 ms 75.6%
  triton_mm_8752 1.7191 ms 74.0%
  triton_mm_8747 2.0869 ms 60.9%
  triton_mm_8748 2.1441 ms 59.3%
SingleProcess AUTOTUNE benchmarking takes 2.3278 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  mm 1.2591 ms 100.0%
  triton_mm_9129 1.2696 ms 99.2%
  triton_mm_9126 1.5471 ms 81.4%
  triton_mm_9122 1.5537 ms 81.0%
  triton_mm_9130 1.5590 ms 80.8%
  triton_mm_9127 1.6656 ms 75.6%
  triton_mm_9123 1.6810 ms 74.9%
  triton_mm_9124 1.6935 ms 74.3%
  triton_mm_9119 2.0723 ms 60.8%
  triton_mm_9120 2.1208 ms 59.4%
SingleProcess AUTOTUNE benchmarking takes 2.2366 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10710 1.2693 ms 100.0%
  mm 1.3040 ms 97.3%
  triton_mm_10707 1.5448 ms 82.2%
  triton_mm_10703 1.5552 ms 81.6%
  triton_mm_10711 1.5558 ms 81.6%
  triton_mm_10708 1.6620 ms 76.4%
  triton_mm_10704 1.6625 ms 76.3%
  triton_mm_10705 1.6875 ms 75.2%
  triton_mm_10700 2.0642 ms 61.5%
  triton_mm_10701 2.1212 ms 59.8%
SingleProcess AUTOTUNE benchmarking takes 2.2482 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3671 ms 100.0%
  mm 0.3766 ms 97.5%
  triton_mm_11504 0.4103 ms 89.5%
  triton_mm_11508 0.4177 ms 87.9%
  triton_mm_11505 0.4437 ms 82.7%
  triton_mm_11506 0.4562 ms 80.5%
  triton_mm_11512 0.4604 ms 79.7%
  triton_mm_11509 0.4629 ms 79.3%
  triton_mm_11501 0.5561 ms 66.0%
  triton_mm_11502 0.5754 ms 63.8%
SingleProcess AUTOTUNE benchmarking takes 2.1482 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3884 ms 100.0%
  mm 0.4012 ms 96.8%
  triton_mm_11523 0.4304 ms 90.3%
  triton_mm_11527 0.4427 ms 87.7%
  triton_mm_11524 0.4708 ms 82.5%
  triton_mm_11525 0.4759 ms 81.6%
  triton_mm_11531 0.4908 ms 79.2%
  triton_mm_11528 0.4916 ms 79.0%
  triton_mm_11520 0.5859 ms 66.3%
  triton_mm_11521 0.6013 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.1550 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8680 0.0315 ms 100.0%
  triton_mm_8681 0.0319 ms 98.9%
  triton_mm_8678 0.0326 ms 96.8%
  triton_mm_8676 0.0326 ms 96.7%
  triton_mm_8675 0.0326 ms 96.6%
  triton_mm_8677 0.0326 ms 96.6%
  triton_mm_8673 0.0344 ms 91.6%
  triton_mm_8684 0.0350 ms 90.1%
  triton_mm_8672 0.0359 ms 87.8%
  triton_mm_8671 0.0361 ms 87.3%
SingleProcess AUTOTUNE benchmarking takes 2.1899 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:22, 119.12s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3951 ms 100.0%
  triton_mm_11523 0.4391 ms 90.0%
  triton_mm_11527 0.4511 ms 87.6%
  mm 0.4556 ms 86.7%
  triton_mm_11524 0.4699 ms 84.1%
  triton_mm_11525 0.4803 ms 82.3%
  triton_mm_11531 0.4891 ms 80.8%
  triton_mm_11528 0.4949 ms 79.8%
  triton_mm_11520 0.5959 ms 66.3%
  triton_mm_11521 0.6120 ms 64.6%
SingleProcess AUTOTUNE benchmarking takes 2.1363 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  bias_addmm 5.4038 ms 100.0%
  triton_mm_11492 5.4552 ms 99.1%
  addmm 6.1154 ms 88.4%
  triton_mm_11489 6.6588 ms 81.2%
  triton_mm_11493 6.6981 ms 80.7%
  triton_mm_11485 6.7675 ms 79.8%
  triton_mm_11490 7.1675 ms 75.4%
  triton_mm_11486 7.2781 ms 74.2%
  triton_mm_11487 7.2898 ms 74.1%
  triton_mm_11482 8.9228 ms 60.6%
SingleProcess AUTOTUNE benchmarking takes 3.0589 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1758 ms 100.0%
  bias_addmm 0.1781 ms 98.7%
  triton_mm_11546 0.1806 ms 97.3%
  triton_mm_11542 0.1807 ms 97.3%
  addmm 0.1842 ms 95.4%
  triton_mm_11543 0.1933 ms 90.9%
  triton_mm_11547 0.1950 ms 90.1%
  triton_mm_11544 0.1965 ms 89.4%
  triton_mm_11550 0.2135 ms 82.3%
  triton_mm_11540 0.2359 ms 74.5%
SingleProcess AUTOTUNE benchmarking takes 2.2106 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:22, 119.14s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3964 ms 100.0%
  triton_mm_11523 0.4389 ms 90.3%
  triton_mm_11527 0.4525 ms 87.6%
  mm 0.4564 ms 86.9%
  triton_mm_11524 0.4717 ms 84.0%
  triton_mm_11525 0.4817 ms 82.3%
  triton_mm_11531 0.4909 ms 80.8%
  triton_mm_11528 0.4972 ms 79.7%
  triton_mm_11520 0.5961 ms 66.5%
  triton_mm_11521 0.6189 ms 64.1%
SingleProcess AUTOTUNE benchmarking takes 2.0621 seconds and 0.0009 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:23, 119.14s/it]AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  triton_mm_8702 0.9787 ms 100.0%
  bias_addmm 1.0817 ms 90.5%
  triton_mm_8699 1.1650 ms 84.0%
  triton_mm_8695 1.1760 ms 83.2%
  triton_mm_8703 1.2169 ms 80.4%
  addmm 1.2275 ms 79.7%
  triton_mm_8700 1.2682 ms 77.2%
  triton_mm_8696 1.2687 ms 77.1%
  triton_mm_8697 1.2844 ms 76.2%
  triton_mm_8692 1.5705 ms 62.3%
SingleProcess AUTOTUNE benchmarking takes 2.4438 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_8721 0.3302 ms 100.0%
  mm 0.3398 ms 97.2%
  triton_mm_8714 0.3952 ms 83.6%
  triton_mm_8718 0.4021 ms 82.1%
  triton_mm_8722 0.4050 ms 81.5%
  triton_mm_8715 0.4309 ms 76.6%
  triton_mm_8716 0.4311 ms 76.6%
  triton_mm_8719 0.4372 ms 75.5%
  triton_mm_8711 0.5217 ms 63.3%
  triton_mm_8712 0.5417 ms 61.0%
SingleProcess AUTOTUNE benchmarking takes 2.1575 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_8776 1.2636 ms 100.0%
  mm 1.2703 ms 99.5%
  triton_mm_8777 1.4850 ms 85.1%
  triton_mm_8769 1.6033 ms 78.8%
  triton_mm_8773 1.6367 ms 77.2%
  triton_mm_8771 1.6790 ms 75.3%
  triton_mm_8770 1.6820 ms 75.1%
  triton_mm_8774 1.6941 ms 74.6%
  triton_mm_8766 2.1222 ms 59.5%
  triton_mm_8767 2.1593 ms 58.5%
SingleProcess AUTOTUNE benchmarking takes 2.3370 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x1920)
  triton_mm_9837 0.3308 ms 100.0%
  mm 0.3382 ms 97.8%
  triton_mm_9830 0.3948 ms 83.8%
  triton_mm_9834 0.4008 ms 82.5%
  triton_mm_9838 0.4051 ms 81.6%
  triton_mm_9831 0.4314 ms 76.7%
  triton_mm_9832 0.4324 ms 76.5%
  triton_mm_9835 0.4370 ms 75.7%
  triton_mm_9827 0.5243 ms 63.1%
  triton_mm_9828 0.5444 ms 60.8%
SingleProcess AUTOTUNE benchmarking takes 2.0692 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x7680, 7680x1920)
  triton_mm_9892 1.2706 ms 100.0%
  mm 1.2709 ms 100.0%
  triton_mm_9893 1.5030 ms 84.5%
  triton_mm_9885 1.6126 ms 78.8%
  triton_mm_9889 1.6364 ms 77.6%
  triton_mm_9887 1.6794 ms 75.7%
  triton_mm_9886 1.6817 ms 75.6%
  triton_mm_9890 1.6935 ms 75.0%
  triton_mm_9882 2.1388 ms 59.4%
  triton_mm_9883 2.1593 ms 58.8%
SingleProcess AUTOTUNE benchmarking takes 2.2587 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x5760, 5148x1920, 1920x5760)
  bias_addmm 0.9781 ms 100.0%
  triton_mm_9911 0.9800 ms 99.8%
  addmm 1.1059 ms 88.4%
  triton_mm_9908 1.1666 ms 83.8%
  triton_mm_9904 1.1771 ms 83.1%
  triton_mm_9912 1.2328 ms 79.3%
  triton_mm_9909 1.2806 ms 76.4%
  triton_mm_9905 1.2815 ms 76.3%
  triton_mm_9906 1.3024 ms 75.1%
  triton_mm_9901 1.5872 ms 61.6%
SingleProcess AUTOTUNE benchmarking takes 2.3245 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(18x1920, 1920x3840)
  triton_mm_11423 0.0383 ms 100.0%
  triton_mm_11424 0.0393 ms 97.6%
  triton_mm_11432 0.0400 ms 95.8%
  triton_mm_11428 0.0401 ms 95.5%
  triton_mm_11437 0.0409 ms 93.8%
  triton_mm_11425 0.0410 ms 93.4%
  triton_mm_11422 0.0411 ms 93.4%
  triton_mm_11435 0.0419 ms 91.5%
  triton_mm_11431 0.0434 ms 88.3%
  mm 0.0447 ms 85.8%
SingleProcess AUTOTUNE benchmarking takes 1.9903 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x32768, 5148x1920, 1920x32768)
  triton_mm_11492 5.3877 ms 100.0%
  bias_addmm 5.3991 ms 99.8%
  addmm 6.0472 ms 89.1%
  triton_mm_11489 6.5931 ms 81.7%
  triton_mm_11493 6.6920 ms 80.5%
  triton_mm_11485 6.7012 ms 80.4%
  triton_mm_11490 7.1882 ms 75.0%
  triton_mm_11486 7.2112 ms 74.7%
  triton_mm_11487 7.2335 ms 74.5%
  triton_mm_11482 8.9467 ms 60.2%
SingleProcess AUTOTUNE benchmarking takes 3.0844 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  triton_mm_11549 0.1763 ms 100.0%
  bias_addmm 0.1776 ms 99.2%
  triton_mm_11542 0.1795 ms 98.2%
  triton_mm_11546 0.1798 ms 98.0%
  addmm 0.1840 ms 95.8%
  triton_mm_11543 0.1925 ms 91.6%
  triton_mm_11547 0.1943 ms 90.7%
  triton_mm_11544 0.1951 ms 90.4%
  triton_mm_11550 0.2144 ms 82.2%
  triton_mm_11540 0.2363 ms 74.6%
SingleProcess AUTOTUNE benchmarking takes 2.2180 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:59<1:25:27, 119.24s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:53, 51.28s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:53, 51.28s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:53, 51.28s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:53, 51.28s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:53, 51.28s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:53, 51.29s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:02<35:54, 51.29s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [02:03<35:55, 51.33s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:27, 28.48s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [02:04<19:28, 28.50s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:47, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:48, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:48, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:48, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:47, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:47, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:48, 17.70s/it]
Validation Progress:   9%|9         | 4/44 [02:05<11:48, 17.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.72s/it]
Validation Progress:  11%|#1        | 5/44 [02:06<07:37, 11.74s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.12s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.12s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.13s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.13s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.12s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.12s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:08,  8.13s/it]
Validation Progress:  14%|#3        | 6/44 [02:07<05:09,  8.13s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:08<03:36,  5.86s/it]
Validation Progress:  16%|#5        | 7/44 [02:09<03:36,  5.86s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.37s/it]
Validation Progress:  18%|#8        | 8/44 [02:10<02:37,  4.38s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  20%|##        | 9/44 [02:11<01:58,  3.40s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:33,  2.74s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:33,  2.74s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:32,  2.74s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:33,  2.74s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:33,  2.74s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:33,  2.74s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:34,  2.77s/it]
Validation Progress:  23%|##2       | 10/44 [02:12<01:33,  2.74s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.29s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.27s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.28s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.28s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.28s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.28s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:15,  2.29s/it]
Validation Progress:  25%|##5       | 11/44 [02:13<01:14,  2.25s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.96s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.97s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.96s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.96s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.96s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.96s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.94s/it]
Validation Progress:  27%|##7       | 12/44 [02:15<01:02,  1.96s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.76s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.75s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.75s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:53,  1.74s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.75s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.76s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.76s/it]
Validation Progress:  30%|##9       | 13/44 [02:16<00:54,  1.75s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.60s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.60s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.60s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.60s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.60s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:48,  1.61s/it]
Validation Progress:  32%|###1      | 14/44 [02:17<00:47,  1.59s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.48s/it]
Validation Progress:  34%|###4      | 15/44 [02:18<00:43,  1.49s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.41s/it]
Validation Progress:  36%|###6      | 16/44 [02:20<00:39,  1.42s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:21<00:37,  1.39s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  41%|####      | 18/44 [02:22<00:35,  1.36s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:24<00:33,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  45%|####5     | 20/44 [02:25<00:31,  1.30s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.27s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:26<00:29,  1.27s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  50%|#####     | 22/44 [02:27<00:27,  1.25s/it]
Validation Progress:  52%|#####2    | 23/44 [02:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:28<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:29<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:29<00:26,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:25,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:25,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:30<00:26,  1.30s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:31<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:32<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:34<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.26s/it]
Validation Progress:  64%|######3   | 28/44 [02:35<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.31s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.31s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.31s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [02:36<00:19,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [02:38<00:18,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [02:39<00:17,  1.32s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [02:40<00:15,  1.33s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:42<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  77%|#######7  | 34/44 [02:43<00:12,  1.27s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  80%|#######9  | 35/44 [02:44<00:11,  1.29s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:45<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  84%|########4 | 37/44 [02:47<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.33s/it]
Validation Progress:  86%|########6 | 38/44 [02:48<00:07,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [02:50<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [02:51<00:05,  1.35s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:52<00:04,  1.34s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  95%|#########5| 42/44 [02:54<00:02,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:55<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  1.20s/it]
Validation Progress: 100%|##########| 44/44 [02:56<00:00,  4.01s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.90s/it]
100%|##########| 1/1 [00:03<00:00,  3.90s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|          | 1/157 [00:01<03:10,  1.22s/it]
  2%|1         | 3/157 [00:01<00:55,  2.77it/s]
  3%|3         | 5/157 [00:01<00:31,  4.84it/s]
  4%|4         | 7/157 [00:01<00:21,  6.86it/s]
  6%|5         | 9/157 [00:01<00:17,  8.67it/s]
  7%|7         | 11/157 [00:01<00:14, 10.39it/s]
  8%|8         | 13/157 [00:01<00:12, 11.76it/s]
 10%|9         | 15/157 [00:02<00:10, 12.98it/s]
 11%|#         | 17/157 [00:02<00:09, 14.06it/s]
 12%|#2        | 19/157 [00:02<00:09, 14.75it/s]
 13%|#3        | 21/157 [00:02<00:09, 14.88it/s]
 15%|#4        | 23/157 [00:02<00:08, 15.47it/s]
 16%|#5        | 25/157 [00:02<00:08, 15.73it/s]
 17%|#7        | 27/157 [00:02<00:08, 16.14it/s]
 18%|#8        | 29/157 [00:02<00:07, 16.45it/s]
 20%|#9        | 31/157 [00:03<00:07, 16.64it/s]
 21%|##1       | 33/157 [00:03<00:07, 16.69it/s]
 22%|##2       | 35/157 [00:03<00:07, 16.72it/s]
 24%|##3       | 37/157 [00:03<00:07, 15.99it/s]
 25%|##4       | 39/157 [00:03<00:07, 16.29it/s]
 26%|##6       | 41/157 [00:03<00:07, 16.48it/s]
 27%|##7       | 43/157 [00:03<00:06, 16.70it/s]
 29%|##8       | 45/157 [00:03<00:06, 16.78it/s]
 30%|##9       | 47/157 [00:04<00:06, 16.66it/s]
 31%|###1      | 49/157 [00:04<00:06, 16.67it/s]
 32%|###2      | 51/157 [00:04<00:06, 16.74it/s]
 34%|###3      | 53/157 [00:04<00:06, 16.87it/s]
 35%|###5      | 55/157 [00:04<00:06, 16.87it/s]
 36%|###6      | 57/157 [00:04<00:05, 16.93it/s]
 38%|###7      | 59/157 [00:04<00:05, 16.89it/s]
 39%|###8      | 61/157 [00:04<00:05, 17.02it/s]
 40%|####      | 63/157 [00:04<00:05, 17.08it/s]
 41%|####1     | 65/157 [00:05<00:05, 17.11it/s]
 43%|####2     | 67/157 [00:05<00:05, 16.86it/s]
 44%|####3     | 69/157 [00:05<00:05, 15.24it/s]
 45%|####5     | 71/157 [00:05<00:05, 15.76it/s]
 46%|####6     | 73/157 [00:05<00:05, 16.02it/s]
 48%|####7     | 75/157 [00:05<00:05, 16.01it/s]
 49%|####9     | 77/157 [00:05<00:04, 16.26it/s]
 50%|#####     | 79/157 [00:05<00:04, 16.48it/s]
 52%|#####1    | 81/157 [00:06<00:04, 16.74it/s]
 53%|#####2    | 83/157 [00:06<00:04, 16.86it/s]
 54%|#####4    | 85/157 [00:06<00:04, 16.94it/s]
 55%|#####5    | 87/157 [00:06<00:04, 17.02it/s]
 57%|#####6    | 89/157 [00:06<00:04, 16.99it/s]
 58%|#####7    | 91/157 [00:06<00:03, 16.81it/s]
 59%|#####9    | 93/157 [00:06<00:03, 16.96it/s]
 61%|######    | 95/157 [00:06<00:03, 17.01it/s]
 62%|######1   | 97/157 [00:07<00:03, 16.99it/s]
 63%|######3   | 99/157 [00:07<00:03, 16.99it/s]
 64%|######4   | 101/157 [00:07<00:03, 16.92it/s]
 66%|######5   | 103/157 [00:07<00:03, 16.98it/s]
 67%|######6   | 105/157 [00:07<00:03, 16.99it/s]
 68%|######8   | 107/157 [00:07<00:02, 17.07it/s]
 69%|######9   | 109/157 [00:07<00:02, 17.10it/s]
 71%|#######   | 111/157 [00:07<00:02, 17.00it/s]
 72%|#######1  | 113/157 [00:07<00:02, 16.96it/s]
 73%|#######3  | 115/157 [00:08<00:02, 16.85it/s]
 75%|#######4  | 117/157 [00:08<00:02, 16.57it/s]
 76%|#######5  | 119/157 [00:08<00:02, 16.70it/s]
 77%|#######7  | 121/157 [00:08<00:02, 16.79it/s]
 78%|#######8  | 123/157 [00:08<00:02, 16.56it/s]
 80%|#######9  | 125/157 [00:08<00:01, 16.37it/s]
 81%|########  | 127/157 [00:08<00:01, 16.24it/s]
 82%|########2 | 129/157 [00:08<00:01, 16.24it/s]
 83%|########3 | 131/157 [00:09<00:01, 16.21it/s]
 85%|########4 | 133/157 [00:09<00:01, 13.59it/s]
 86%|########5 | 135/157 [00:09<00:01, 14.31it/s]
 87%|########7 | 137/157 [00:09<00:01, 14.85it/s]
 89%|########8 | 139/157 [00:09<00:01, 15.24it/s]
 90%|########9 | 141/157 [00:09<00:01, 15.34it/s]
 91%|#########1| 143/157 [00:09<00:00, 15.59it/s]
 92%|#########2| 145/157 [00:09<00:00, 15.91it/s]
 94%|#########3| 147/157 [00:10<00:00, 16.00it/s]
 95%|#########4| 149/157 [00:10<00:00, 16.28it/s]
 96%|#########6| 151/157 [00:10<00:00, 16.37it/s]
 97%|#########7| 153/157 [00:10<00:00, 16.45it/s]
 99%|#########8| 155/157 [00:10<00:00, 16.44it/s]
100%|##########| 157/157 [00:11<00:00,  5.80it/s]
100%|##########| 157/157 [00:11<00:00, 13.70it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:43, 17.90it/s]
  1%|          | 4/781 [00:00<00:42, 18.11it/s]
  1%|          | 6/781 [00:00<00:42, 18.16it/s]
  1%|1         | 8/781 [00:00<00:42, 18.23it/s]
  1%|1         | 10/781 [00:00<00:42, 18.34it/s]
  2%|1         | 12/781 [00:00<00:42, 18.24it/s]
  2%|1         | 14/781 [00:00<00:43, 17.78it/s]
  2%|2         | 16/781 [00:00<00:42, 18.17it/s]
  2%|2         | 18/781 [00:00<00:41, 18.43it/s]
  3%|2         | 20/781 [00:01<00:41, 18.46it/s]
  3%|2         | 22/781 [00:01<00:40, 18.55it/s]
  3%|3         | 24/781 [00:01<00:41, 18.31it/s]
  3%|3         | 26/781 [00:01<00:40, 18.52it/s]
  4%|3         | 28/781 [00:01<00:40, 18.68it/s]
  4%|3         | 30/781 [00:01<00:40, 18.73it/s]
  4%|4         | 32/781 [00:01<00:39, 18.83it/s]
  4%|4         | 34/781 [00:01<00:39, 18.75it/s]
  5%|4         | 36/781 [00:01<00:39, 18.76it/s]
  5%|4         | 38/781 [00:02<00:39, 18.84it/s]
  5%|5         | 40/781 [00:02<00:39, 18.82it/s]
  5%|5         | 42/781 [00:02<00:39, 18.77it/s]
  6%|5         | 44/781 [00:02<00:39, 18.87it/s]
  6%|5         | 46/781 [00:02<00:38, 18.87it/s]
  6%|6         | 48/781 [00:02<00:38, 18.99it/s]
  6%|6         | 50/781 [00:02<00:38, 18.92it/s]
  7%|6         | 52/781 [00:02<00:38, 19.00it/s]
  7%|6         | 54/781 [00:02<00:38, 19.07it/s]
  7%|7         | 56/781 [00:03<00:38, 18.97it/s]
  7%|7         | 58/781 [00:03<00:38, 19.00it/s]
  8%|7         | 60/781 [00:03<00:38, 18.89it/s]
  8%|7         | 62/781 [00:03<00:38, 18.91it/s]
  8%|8         | 64/781 [00:03<00:37, 18.92it/s]
  8%|8         | 66/781 [00:03<00:37, 18.89it/s]
  9%|8         | 68/781 [00:03<00:37, 18.95it/s]
  9%|8         | 70/781 [00:03<00:37, 18.92it/s]
  9%|9         | 72/781 [00:03<00:37, 18.78it/s]
  9%|9         | 74/781 [00:03<00:37, 18.62it/s]
 10%|9         | 76/781 [00:04<00:37, 18.75it/s]
 10%|9         | 78/781 [00:04<00:37, 18.88it/s]
 10%|#         | 80/781 [00:04<00:37, 18.66it/s]
 10%|#         | 82/781 [00:04<00:37, 18.69it/s]
 11%|#         | 84/781 [00:04<00:37, 18.66it/s]
 11%|#1        | 86/781 [00:04<00:37, 18.67it/s]
 11%|#1        | 88/781 [00:04<00:37, 18.62it/s]
 12%|#1        | 90/781 [00:04<00:37, 18.57it/s]
 12%|#1        | 92/781 [00:04<00:37, 18.47it/s]
 12%|#2        | 94/781 [00:05<00:37, 18.34it/s]
 12%|#2        | 96/781 [00:05<00:37, 18.40it/s]
 13%|#2        | 98/781 [00:05<00:37, 18.34it/s]
 13%|#2        | 100/781 [00:05<00:36, 18.49it/s]
 13%|#3        | 102/781 [00:05<00:36, 18.43it/s]
 13%|#3        | 104/781 [00:05<00:36, 18.39it/s]
 14%|#3        | 106/781 [00:05<00:36, 18.41it/s]
 14%|#3        | 108/781 [00:05<00:36, 18.39it/s]
 14%|#4        | 110/781 [00:05<00:37, 18.11it/s]
 14%|#4        | 112/781 [00:06<00:36, 18.09it/s]
 15%|#4        | 114/781 [00:06<00:36, 18.39it/s]
 15%|#4        | 116/781 [00:06<00:36, 18.32it/s]
 15%|#5        | 118/781 [00:06<00:36, 18.24it/s]
 15%|#5        | 120/781 [00:06<00:35, 18.48it/s]
 16%|#5        | 122/781 [00:06<00:35, 18.67it/s]
 16%|#5        | 124/781 [00:06<00:34, 18.95it/s]
 16%|#6        | 126/781 [00:06<00:34, 19.10it/s]
 16%|#6        | 128/781 [00:06<00:33, 19.26it/s]
 17%|#6        | 130/781 [00:06<00:33, 19.38it/s]
 17%|#6        | 132/781 [00:07<00:33, 19.44it/s]
 17%|#7        | 134/781 [00:07<00:33, 19.43it/s]
 17%|#7        | 136/781 [00:07<00:33, 19.28it/s]
 18%|#7        | 138/781 [00:07<00:33, 19.47it/s]
 18%|#7        | 140/781 [00:07<00:32, 19.47it/s]
 18%|#8        | 142/781 [00:07<00:32, 19.58it/s]
 18%|#8        | 144/781 [00:07<00:32, 19.54it/s]
 19%|#8        | 146/781 [00:07<00:32, 19.63it/s]
 19%|#8        | 148/781 [00:07<00:32, 19.63it/s]
 19%|#9        | 150/781 [00:07<00:32, 19.61it/s]
 19%|#9        | 152/781 [00:08<00:32, 19.63it/s]
 20%|#9        | 154/781 [00:08<00:32, 19.55it/s]
 20%|#9        | 156/781 [00:08<00:32, 19.52it/s]
 20%|##        | 158/781 [00:08<00:32, 19.36it/s]
 20%|##        | 160/781 [00:08<00:32, 19.32it/s]
 21%|##        | 162/781 [00:08<00:31, 19.45it/s]
 21%|##        | 164/781 [00:08<00:32, 19.16it/s]
 21%|##1       | 166/781 [00:08<00:32, 18.81it/s]
 22%|##1       | 168/781 [00:08<00:32, 18.77it/s]
 22%|##1       | 170/781 [00:09<00:32, 18.79it/s]
 22%|##2       | 172/781 [00:09<00:32, 18.74it/s]
 22%|##2       | 174/781 [00:09<00:32, 18.78it/s]
 23%|##2       | 176/781 [00:09<00:32, 18.89it/s]
 23%|##2       | 178/781 [00:09<00:31, 18.90it/s]
 23%|##3       | 180/781 [00:09<00:31, 18.99it/s]
 23%|##3       | 182/781 [00:09<00:32, 18.64it/s]
 24%|##3       | 184/781 [00:09<00:31, 18.75it/s]
 24%|##3       | 186/781 [00:09<00:31, 19.01it/s]
 24%|##4       | 188/781 [00:09<00:31, 18.97it/s]
 24%|##4       | 190/781 [00:10<00:30, 19.15it/s]
 25%|##4       | 192/781 [00:10<00:30, 19.27it/s]
 25%|##4       | 194/781 [00:10<00:30, 19.17it/s]
 25%|##5       | 196/781 [00:10<00:30, 19.07it/s]
 25%|##5       | 198/781 [00:10<00:30, 18.99it/s]
 26%|##5       | 200/781 [00:10<00:30, 18.88it/s]
 26%|##5       | 202/781 [00:10<00:30, 19.03it/s]
 26%|##6       | 204/781 [00:10<00:29, 19.28it/s]
 26%|##6       | 206/781 [00:10<00:29, 19.34it/s]
 27%|##6       | 208/781 [00:11<00:29, 19.50it/s]
 27%|##6       | 210/781 [00:11<00:29, 19.38it/s]
 27%|##7       | 212/781 [00:11<00:29, 19.52it/s]
 27%|##7       | 214/781 [00:11<00:28, 19.56it/s]
 28%|##7       | 216/781 [00:11<00:28, 19.59it/s]
 28%|##7       | 218/781 [00:11<00:28, 19.64it/s]
 28%|##8       | 220/781 [00:11<00:28, 19.68it/s]
 28%|##8       | 222/781 [00:11<00:28, 19.76it/s]
 29%|##8       | 224/781 [00:11<00:28, 19.80it/s]
 29%|##8       | 226/781 [00:11<00:28, 19.65it/s]
 29%|##9       | 228/781 [00:12<00:28, 19.20it/s]
 29%|##9       | 230/781 [00:12<00:28, 19.17it/s]
 30%|##9       | 232/781 [00:12<00:28, 19.39it/s]
 30%|##9       | 234/781 [00:12<00:28, 19.47it/s]
 30%|###       | 236/781 [00:12<00:27, 19.50it/s]
 30%|###       | 238/781 [00:12<00:28, 19.22it/s]
 31%|###       | 240/781 [00:12<00:27, 19.39it/s]
 31%|###       | 242/781 [00:12<00:27, 19.51it/s]
 31%|###1      | 244/781 [00:12<00:27, 19.53it/s]
 31%|###1      | 246/781 [00:12<00:27, 19.57it/s]
 32%|###1      | 248/781 [00:13<00:27, 19.54it/s]
 32%|###2      | 250/781 [00:13<00:27, 19.39it/s]
 32%|###2      | 252/781 [00:13<00:27, 19.42it/s]
 33%|###2      | 254/781 [00:13<00:27, 19.39it/s]
 33%|###2      | 256/781 [00:13<00:27, 19.34it/s]
 33%|###3      | 258/781 [00:13<00:27, 19.32it/s]
 33%|###3      | 260/781 [00:13<00:27, 19.20it/s]
 34%|###3      | 262/781 [00:13<00:38, 13.43it/s]
 34%|###3      | 264/781 [00:14<00:35, 14.73it/s]
 34%|###4      | 266/781 [00:14<00:32, 15.83it/s]
 34%|###4      | 268/781 [00:14<00:30, 16.79it/s]
 35%|###4      | 270/781 [00:14<00:29, 17.50it/s]
 35%|###4      | 272/781 [00:14<00:28, 18.11it/s]
 35%|###5      | 274/781 [00:14<00:27, 18.58it/s]
 35%|###5      | 276/781 [00:14<00:26, 18.89it/s]
 36%|###5      | 278/781 [00:14<00:26, 19.16it/s]
 36%|###5      | 280/781 [00:14<00:25, 19.34it/s]
 36%|###6      | 282/781 [00:14<00:25, 19.36it/s]
 36%|###6      | 284/781 [00:15<00:25, 19.41it/s]
 37%|###6      | 286/781 [00:15<00:25, 19.45it/s]
 37%|###6      | 288/781 [00:15<00:25, 19.51it/s]
 37%|###7      | 290/781 [00:15<00:25, 19.54it/s]
 37%|###7      | 292/781 [00:15<00:25, 19.24it/s]
 38%|###7      | 294/781 [00:15<00:25, 19.31it/s]
 38%|###7      | 296/781 [00:15<00:24, 19.48it/s]
 38%|###8      | 298/781 [00:15<00:24, 19.59it/s]
 38%|###8      | 300/781 [00:15<00:24, 19.66it/s]
 39%|###8      | 302/781 [00:16<00:24, 19.43it/s]
 39%|###8      | 304/781 [00:16<00:24, 19.58it/s]
 39%|###9      | 306/781 [00:16<00:24, 19.70it/s]
 39%|###9      | 308/781 [00:16<00:23, 19.76it/s]
 40%|###9      | 310/781 [00:16<00:24, 19.39it/s]
 40%|###9      | 312/781 [00:16<00:24, 19.54it/s]
 40%|####      | 314/781 [00:16<00:23, 19.64it/s]
 40%|####      | 316/781 [00:16<00:23, 19.67it/s]
 41%|####      | 318/781 [00:16<00:23, 19.57it/s]
 41%|####      | 320/781 [00:16<00:23, 19.49it/s]
 41%|####1     | 322/781 [00:17<00:23, 19.44it/s]
 41%|####1     | 324/781 [00:17<00:23, 19.33it/s]
 42%|####1     | 326/781 [00:17<00:23, 19.36it/s]
 42%|####1     | 328/781 [00:17<00:23, 19.53it/s]
 42%|####2     | 330/781 [00:17<00:23, 19.47it/s]
 43%|####2     | 332/781 [00:17<00:23, 19.47it/s]
 43%|####2     | 334/781 [00:17<00:23, 19.41it/s]
 43%|####3     | 336/781 [00:17<00:22, 19.51it/s]
 43%|####3     | 338/781 [00:17<00:22, 19.62it/s]
 44%|####3     | 340/781 [00:17<00:22, 19.48it/s]
 44%|####3     | 342/781 [00:18<00:23, 18.79it/s]
 44%|####4     | 344/781 [00:18<00:23, 18.92it/s]
 44%|####4     | 346/781 [00:18<00:22, 19.03it/s]
 45%|####4     | 348/781 [00:18<00:22, 19.21it/s]
 45%|####4     | 350/781 [00:18<00:22, 19.39it/s]
 45%|####5     | 352/781 [00:18<00:22, 19.06it/s]
 45%|####5     | 354/781 [00:18<00:22, 19.19it/s]
 46%|####5     | 356/781 [00:18<00:22, 19.14it/s]
 46%|####5     | 358/781 [00:18<00:22, 18.95it/s]
 46%|####6     | 360/781 [00:19<00:22, 18.96it/s]
 46%|####6     | 362/781 [00:19<00:21, 19.08it/s]
 47%|####6     | 364/781 [00:19<00:21, 19.07it/s]
 47%|####6     | 366/781 [00:19<00:21, 19.04it/s]
 47%|####7     | 368/781 [00:19<00:21, 18.95it/s]
 47%|####7     | 370/781 [00:19<00:21, 19.13it/s]
 48%|####7     | 372/781 [00:19<00:21, 19.31it/s]
 48%|####7     | 374/781 [00:19<00:20, 19.44it/s]
 48%|####8     | 376/781 [00:19<00:20, 19.48it/s]
 48%|####8     | 378/781 [00:19<00:20, 19.44it/s]
 49%|####8     | 380/781 [00:20<00:21, 19.02it/s]
 49%|####8     | 382/781 [00:20<00:20, 19.07it/s]
 49%|####9     | 384/781 [00:20<00:20, 19.05it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.05it/s]
 50%|####9     | 388/781 [00:20<00:20, 19.09it/s]
 50%|####9     | 390/781 [00:20<00:20, 19.22it/s]
 50%|#####     | 392/781 [00:20<00:20, 19.37it/s]
 50%|#####     | 394/781 [00:20<00:19, 19.54it/s]
 51%|#####     | 396/781 [00:20<00:19, 19.64it/s]
 51%|#####     | 398/781 [00:20<00:19, 19.68it/s]
 51%|#####1    | 400/781 [00:21<00:19, 19.73it/s]
 51%|#####1    | 402/781 [00:21<00:19, 19.75it/s]
 52%|#####1    | 404/781 [00:21<00:19, 19.79it/s]
 52%|#####1    | 406/781 [00:21<00:19, 19.43it/s]
 52%|#####2    | 408/781 [00:21<00:19, 19.22it/s]
 52%|#####2    | 410/781 [00:21<00:19, 19.12it/s]
 53%|#####2    | 412/781 [00:21<00:19, 19.07it/s]
 53%|#####3    | 414/781 [00:21<00:19, 19.20it/s]
 53%|#####3    | 416/781 [00:21<00:19, 18.82it/s]
 54%|#####3    | 418/781 [00:22<00:19, 18.76it/s]
 54%|#####3    | 420/781 [00:22<00:19, 18.61it/s]
 54%|#####4    | 422/781 [00:22<00:19, 18.52it/s]
 54%|#####4    | 424/781 [00:22<00:19, 18.54it/s]
 55%|#####4    | 426/781 [00:22<00:19, 18.56it/s]
 55%|#####4    | 428/781 [00:22<00:18, 18.68it/s]
 55%|#####5    | 430/781 [00:22<00:18, 18.77it/s]
 55%|#####5    | 432/781 [00:22<00:18, 19.01it/s]
 56%|#####5    | 434/781 [00:22<00:18, 19.18it/s]
 56%|#####5    | 436/781 [00:22<00:17, 19.33it/s]
 56%|#####6    | 438/781 [00:23<00:17, 19.42it/s]
 56%|#####6    | 440/781 [00:23<00:17, 19.54it/s]
 57%|#####6    | 442/781 [00:23<00:17, 19.62it/s]
 57%|#####6    | 444/781 [00:23<00:17, 19.65it/s]
 57%|#####7    | 446/781 [00:23<00:17, 19.66it/s]
 57%|#####7    | 448/781 [00:23<00:17, 19.47it/s]
 58%|#####7    | 450/781 [00:23<00:17, 19.44it/s]
 58%|#####7    | 452/781 [00:23<00:17, 19.33it/s]
 58%|#####8    | 454/781 [00:23<00:17, 18.92it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.97it/s]
 59%|#####8    | 458/781 [00:24<00:16, 19.02it/s]
 59%|#####8    | 460/781 [00:24<00:17, 18.42it/s]
 59%|#####9    | 462/781 [00:24<00:17, 18.50it/s]
 59%|#####9    | 464/781 [00:24<00:17, 18.43it/s]
 60%|#####9    | 466/781 [00:24<00:16, 18.63it/s]
 60%|#####9    | 468/781 [00:24<00:16, 18.62it/s]
 60%|######    | 470/781 [00:24<00:16, 18.67it/s]
 60%|######    | 472/781 [00:24<00:16, 18.69it/s]
 61%|######    | 474/781 [00:24<00:16, 18.69it/s]
 61%|######    | 476/781 [00:25<00:16, 18.75it/s]
 61%|######1   | 478/781 [00:25<00:16, 18.72it/s]
 61%|######1   | 480/781 [00:25<00:16, 18.79it/s]
 62%|######1   | 482/781 [00:25<00:15, 18.84it/s]
 62%|######1   | 484/781 [00:25<00:15, 18.87it/s]
 62%|######2   | 486/781 [00:25<00:15, 18.95it/s]
 62%|######2   | 488/781 [00:25<00:15, 19.00it/s]
 63%|######2   | 490/781 [00:25<00:15, 18.87it/s]
 63%|######2   | 492/781 [00:25<00:15, 18.90it/s]
 63%|######3   | 494/781 [00:26<00:15, 18.93it/s]
 64%|######3   | 496/781 [00:26<00:14, 19.09it/s]
 64%|######3   | 498/781 [00:26<00:14, 18.93it/s]
 64%|######4   | 500/781 [00:26<00:14, 18.84it/s]
 64%|######4   | 502/781 [00:26<00:14, 18.76it/s]
 65%|######4   | 504/781 [00:26<00:14, 18.78it/s]
 65%|######4   | 506/781 [00:26<00:14, 18.85it/s]
 65%|######5   | 508/781 [00:26<00:14, 18.83it/s]
 65%|######5   | 510/781 [00:26<00:14, 18.90it/s]
 66%|######5   | 512/781 [00:26<00:14, 18.80it/s]
 66%|######5   | 514/781 [00:27<00:14, 18.75it/s]
 66%|######6   | 516/781 [00:27<00:14, 18.68it/s]
 66%|######6   | 518/781 [00:27<00:13, 18.85it/s]
 67%|######6   | 520/781 [00:27<00:24, 10.49it/s]
 67%|######6   | 522/781 [00:27<00:21, 12.15it/s]
 67%|######7   | 524/781 [00:27<00:18, 13.58it/s]
 67%|######7   | 526/781 [00:28<00:17, 14.85it/s]
 68%|######7   | 528/781 [00:28<00:15, 15.86it/s]
 68%|######7   | 530/781 [00:28<00:15, 16.59it/s]
 68%|######8   | 532/781 [00:28<00:14, 17.19it/s]
 68%|######8   | 534/781 [00:28<00:14, 17.57it/s]
 69%|######8   | 536/781 [00:28<00:13, 17.84it/s]
 69%|######8   | 538/781 [00:28<00:13, 18.10it/s]
 69%|######9   | 540/781 [00:28<00:13, 18.31it/s]
 69%|######9   | 542/781 [00:28<00:13, 18.27it/s]
 70%|######9   | 544/781 [00:28<00:12, 18.35it/s]
 70%|######9   | 546/781 [00:29<00:12, 18.51it/s]
 70%|#######   | 548/781 [00:29<00:12, 18.50it/s]
 70%|#######   | 550/781 [00:29<00:12, 18.47it/s]
 71%|#######   | 552/781 [00:29<00:12, 18.48it/s]
 71%|#######   | 554/781 [00:29<00:12, 18.51it/s]
 71%|#######1  | 556/781 [00:29<00:12, 18.54it/s]
 71%|#######1  | 558/781 [00:29<00:11, 18.80it/s]
 72%|#######1  | 560/781 [00:29<00:12, 17.89it/s]
 72%|#######1  | 562/781 [00:29<00:12, 18.21it/s]
 72%|#######2  | 564/781 [00:30<00:11, 18.35it/s]
 72%|#######2  | 566/781 [00:30<00:11, 18.47it/s]
 73%|#######2  | 568/781 [00:30<00:11, 18.27it/s]
 73%|#######2  | 570/781 [00:30<00:11, 18.44it/s]
 73%|#######3  | 572/781 [00:30<00:11, 18.60it/s]
 73%|#######3  | 574/781 [00:30<00:11, 18.72it/s]
 74%|#######3  | 576/781 [00:30<00:10, 18.69it/s]
 74%|#######4  | 578/781 [00:30<00:10, 18.77it/s]
 74%|#######4  | 580/781 [00:30<00:10, 18.90it/s]
 75%|#######4  | 582/781 [00:31<00:10, 18.94it/s]
 75%|#######4  | 584/781 [00:31<00:10, 19.13it/s]
 75%|#######5  | 586/781 [00:31<00:10, 19.35it/s]
 75%|#######5  | 588/781 [00:31<00:09, 19.51it/s]
 76%|#######5  | 590/781 [00:31<00:09, 19.48it/s]
 76%|#######5  | 592/781 [00:31<00:09, 19.55it/s]
 76%|#######6  | 594/781 [00:31<00:09, 19.59it/s]
 76%|#######6  | 596/781 [00:31<00:09, 19.36it/s]
 77%|#######6  | 598/781 [00:31<00:09, 19.30it/s]
 77%|#######6  | 600/781 [00:31<00:09, 19.36it/s]
 77%|#######7  | 602/781 [00:32<00:09, 19.50it/s]
 77%|#######7  | 604/781 [00:32<00:09, 19.37it/s]
 78%|#######7  | 606/781 [00:32<00:09, 19.29it/s]
 78%|#######7  | 608/781 [00:32<00:08, 19.43it/s]
 78%|#######8  | 610/781 [00:32<00:08, 19.50it/s]
 78%|#######8  | 612/781 [00:32<00:08, 19.53it/s]
 79%|#######8  | 614/781 [00:32<00:08, 19.44it/s]
 79%|#######8  | 616/781 [00:32<00:08, 19.29it/s]
 79%|#######9  | 618/781 [00:32<00:08, 19.23it/s]
 79%|#######9  | 620/781 [00:32<00:08, 19.17it/s]
 80%|#######9  | 622/781 [00:33<00:08, 19.18it/s]
 80%|#######9  | 624/781 [00:33<00:08, 19.26it/s]
 80%|########  | 626/781 [00:33<00:08, 19.11it/s]
 80%|########  | 628/781 [00:33<00:07, 19.26it/s]
 81%|########  | 630/781 [00:33<00:07, 19.39it/s]
 81%|########  | 632/781 [00:33<00:07, 19.51it/s]
 81%|########1 | 634/781 [00:33<00:07, 18.68it/s]
 81%|########1 | 636/781 [00:33<00:07, 19.00it/s]
 82%|########1 | 638/781 [00:33<00:07, 19.22it/s]
 82%|########1 | 640/781 [00:34<00:07, 19.39it/s]
 82%|########2 | 642/781 [00:34<00:07, 19.56it/s]
 82%|########2 | 644/781 [00:34<00:07, 19.48it/s]
 83%|########2 | 646/781 [00:34<00:06, 19.54it/s]
 83%|########2 | 648/781 [00:34<00:06, 19.60it/s]
 83%|########3 | 650/781 [00:34<00:06, 19.55it/s]
 83%|########3 | 652/781 [00:34<00:06, 19.63it/s]
 84%|########3 | 654/781 [00:34<00:06, 19.63it/s]
 84%|########3 | 656/781 [00:34<00:06, 19.24it/s]
 84%|########4 | 658/781 [00:34<00:06, 19.08it/s]
 85%|########4 | 660/781 [00:35<00:06, 19.18it/s]
 85%|########4 | 662/781 [00:35<00:06, 19.27it/s]
 85%|########5 | 664/781 [00:35<00:06, 19.46it/s]
 85%|########5 | 666/781 [00:35<00:05, 19.59it/s]
 86%|########5 | 668/781 [00:35<00:05, 19.59it/s]
 86%|########5 | 670/781 [00:35<00:05, 19.56it/s]
 86%|########6 | 672/781 [00:35<00:05, 19.56it/s]
 86%|########6 | 674/781 [00:35<00:05, 19.54it/s]
 87%|########6 | 676/781 [00:35<00:05, 19.65it/s]
 87%|########6 | 678/781 [00:35<00:05, 19.73it/s]
 87%|########7 | 680/781 [00:36<00:05, 19.75it/s]
 87%|########7 | 682/781 [00:36<00:05, 19.73it/s]
 88%|########7 | 684/781 [00:36<00:04, 19.56it/s]
 88%|########7 | 686/781 [00:36<00:04, 19.61it/s]
 88%|########8 | 688/781 [00:36<00:04, 19.68it/s]
 88%|########8 | 690/781 [00:36<00:04, 19.60it/s]
 89%|########8 | 692/781 [00:36<00:04, 19.60it/s]
 89%|########8 | 694/781 [00:36<00:04, 19.48it/s]
 89%|########9 | 696/781 [00:36<00:04, 19.26it/s]
 89%|########9 | 698/781 [00:36<00:04, 19.20it/s]
 90%|########9 | 700/781 [00:37<00:04, 19.04it/s]
 90%|########9 | 702/781 [00:37<00:04, 19.05it/s]
 90%|######### | 704/781 [00:37<00:03, 19.28it/s]
 90%|######### | 706/781 [00:37<00:03, 19.44it/s]
 91%|######### | 708/781 [00:37<00:03, 19.23it/s]
 91%|######### | 710/781 [00:37<00:03, 19.22it/s]
 91%|#########1| 712/781 [00:37<00:03, 19.38it/s]
 91%|#########1| 714/781 [00:37<00:03, 19.44it/s]
 92%|#########1| 716/781 [00:37<00:03, 19.39it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.46it/s]
 92%|#########2| 720/781 [00:38<00:03, 19.53it/s]
 92%|#########2| 722/781 [00:38<00:03, 19.49it/s]
 93%|#########2| 724/781 [00:38<00:02, 19.58it/s]
 93%|#########2| 726/781 [00:38<00:02, 19.67it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.55it/s]
 93%|#########3| 730/781 [00:38<00:02, 19.58it/s]
 94%|#########3| 732/781 [00:38<00:02, 19.43it/s]
 94%|#########3| 734/781 [00:38<00:02, 19.20it/s]
 94%|#########4| 736/781 [00:38<00:02, 18.85it/s]
 94%|#########4| 738/781 [00:39<00:02, 18.96it/s]
 95%|#########4| 740/781 [00:39<00:02, 18.89it/s]
 95%|#########5| 742/781 [00:39<00:02, 18.85it/s]
 95%|#########5| 744/781 [00:39<00:01, 18.93it/s]
 96%|#########5| 746/781 [00:39<00:01, 18.96it/s]
 96%|#########5| 748/781 [00:39<00:01, 18.72it/s]
 96%|#########6| 750/781 [00:39<00:01, 18.87it/s]
 96%|#########6| 752/781 [00:39<00:01, 18.83it/s]
 97%|#########6| 754/781 [00:39<00:01, 18.92it/s]
 97%|#########6| 756/781 [00:40<00:01, 18.84it/s]
 97%|#########7| 758/781 [00:40<00:01, 19.02it/s]
 97%|#########7| 760/781 [00:40<00:01, 18.99it/s]
 98%|#########7| 762/781 [00:40<00:01, 18.94it/s]
 98%|#########7| 764/781 [00:40<00:00, 18.97it/s]
 98%|#########8| 766/781 [00:40<00:00, 19.15it/s]
 98%|#########8| 768/781 [00:40<00:00, 19.20it/s]
 99%|#########8| 770/781 [00:40<00:00, 19.02it/s]
 99%|#########8| 772/781 [00:40<00:00, 19.15it/s]
 99%|#########9| 774/781 [00:40<00:00, 19.20it/s]
 99%|#########9| 776/781 [00:41<00:00, 19.37it/s]
100%|#########9| 778/781 [00:41<00:00, 19.35it/s]
100%|#########9| 780/781 [00:41<00:00, 19.19it/s]
100%|##########| 781/781 [00:41<00:00, 18.90it/s]






=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================







=======================================================   RESTART [11-26 09:50:26]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241125_175353-eyi61gk3
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run vivid-gorge-551
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/eyi61gk3







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================







=======================================================   RESTART [11-26 11:11:13]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler







=======================================================   RESTART [11-26 11:28:09]   =======================================================







=======================================================   RESTART [11-26 11:28:10]   =======================================================







=======================================================   RESTART [11-26 11:28:11]   =======================================================







=======================================================   RESTART [11-26 11:28:11]   =======================================================







=======================================================   RESTART [11-26 11:28:11]   =======================================================







=======================================================   RESTART [11-26 11:28:11]   =======================================================







=======================================================   RESTART [11-26 11:28:11]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================







=======================================================   RESTART [11-26 11:39:09]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241125_194234-fcqajq1p
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run vague-glade-552
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/fcqajq1p
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3894 ms 100.0%
  triton_mm_11523 0.4305 ms 90.5%
  triton_mm_11527 0.4430 ms 87.9%
  mm 0.4562 ms 85.4%
  triton_mm_11524 0.4637 ms 84.0%
  triton_mm_11525 0.4753 ms 81.9%
  triton_mm_11531 0.4854 ms 80.2%
  triton_mm_11528 0.4872 ms 79.9%
  triton_mm_11520 0.5918 ms 65.8%
  triton_mm_11521 0.6131 ms 63.5%
SingleProcess AUTOTUNE benchmarking takes 2.1673 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:33, 109.61s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3905 ms 100.0%
  triton_mm_11523 0.4300 ms 90.8%
  triton_mm_11527 0.4411 ms 88.5%
  mm 0.4558 ms 85.7%
  triton_mm_11524 0.4619 ms 84.5%
  triton_mm_11525 0.4735 ms 82.5%
  triton_mm_11531 0.4856 ms 80.4%
  triton_mm_11528 0.4860 ms 80.3%
  triton_mm_11520 0.5840 ms 66.9%
  triton_mm_11521 0.6059 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.1413 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:33, 109.62s/it]AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8680 0.0317 ms 100.0%
  triton_mm_8681 0.0321 ms 98.9%
  triton_mm_8677 0.0327 ms 97.2%
  triton_mm_8678 0.0327 ms 97.2%
  triton_mm_8675 0.0327 ms 97.1%
  triton_mm_8676 0.0328 ms 96.7%
  triton_mm_8673 0.0346 ms 91.7%
  triton_mm_8672 0.0352 ms 90.1%
  triton_mm_8684 0.0358 ms 88.6%
  triton_mm_8671 0.0362 ms 87.6%
SingleProcess AUTOTUNE benchmarking takes 2.2032 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:33, 109.62s/it]AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3726 ms 100.0%
  triton_mm_11504 0.4140 ms 90.0%
  triton_mm_11508 0.4211 ms 88.5%
  mm 0.4320 ms 86.3%
  triton_mm_11505 0.4476 ms 83.3%
  triton_mm_11506 0.4581 ms 81.3%
  triton_mm_11512 0.4635 ms 80.4%
  triton_mm_11509 0.4670 ms 79.8%
  triton_mm_11501 0.5601 ms 66.5%
  triton_mm_11502 0.5774 ms 64.5%
SingleProcess AUTOTUNE benchmarking takes 2.1580 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3957 ms 100.0%
  mm 0.4098 ms 96.6%
  triton_mm_11523 0.4391 ms 90.1%
  triton_mm_11527 0.4485 ms 88.2%
  triton_mm_11524 0.4706 ms 84.1%
  triton_mm_11525 0.4809 ms 82.3%
  triton_mm_11531 0.4885 ms 81.0%
  triton_mm_11528 0.4954 ms 79.9%
  triton_mm_11520 0.5966 ms 66.3%
  triton_mm_11521 0.6197 ms 63.9%
SingleProcess AUTOTUNE benchmarking takes 2.1466 seconds and 0.0000 seconds precompiling
AUTOTUNE addmm(5148x768, 5148x2048, 2048x768)
  bias_addmm 0.1760 ms 100.0%
  triton_mm_11549 0.1763 ms 99.9%
  triton_mm_11546 0.1787 ms 98.5%
  triton_mm_11542 0.1802 ms 97.7%
  addmm 0.1824 ms 96.5%
  triton_mm_11543 0.1920 ms 91.7%
  triton_mm_11547 0.1939 ms 90.8%
  triton_mm_11544 0.2000 ms 88.0%
  triton_mm_11550 0.2143 ms 82.1%
  triton_mm_11540 0.2358 ms 74.7%
SingleProcess AUTOTUNE benchmarking takes 2.2027 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:33, 109.62s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:33, 109.62s/it]AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_8757 1.2661 ms 100.0%
  mm 1.4435 ms 87.7%
  triton_mm_8758 1.5732 ms 80.5%
  triton_mm_8750 1.5751 ms 80.4%
  triton_mm_8754 1.5791 ms 80.2%
  triton_mm_8751 1.6830 ms 75.2%
  triton_mm_8755 1.6878 ms 75.0%
  triton_mm_8752 1.6909 ms 74.9%
  triton_mm_8747 2.0654 ms 61.3%
  triton_mm_8748 2.1432 ms 59.1%
SingleProcess AUTOTUNE benchmarking takes 2.3199 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x7680)
  triton_mm_10245 1.2657 ms 100.0%
  mm 1.2911 ms 98.0%
  triton_mm_10242 1.5645 ms 80.9%
  triton_mm_10246 1.5788 ms 80.2%
  triton_mm_10238 1.5962 ms 79.3%
  triton_mm_10243 1.6832 ms 75.2%
  triton_mm_10240 1.7145 ms 73.8%
  triton_mm_10239 1.7224 ms 73.5%
  triton_mm_10235 2.0919 ms 60.5%
  triton_mm_10236 2.1687 ms 58.4%
SingleProcess AUTOTUNE benchmarking takes 2.2408 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x1920, 1920x2048)
  triton_mm_11511 0.3709 ms 100.0%
  mm 0.3814 ms 97.2%
  triton_mm_11504 0.4152 ms 89.3%
  triton_mm_11508 0.4218 ms 87.9%
  triton_mm_11505 0.4485 ms 82.7%
  triton_mm_11506 0.4566 ms 81.2%
  triton_mm_11512 0.4619 ms 80.3%
  triton_mm_11509 0.4662 ms 79.6%
  triton_mm_11501 0.5606 ms 66.2%
  triton_mm_11502 0.5796 ms 64.0%
SingleProcess AUTOTUNE benchmarking takes 2.1674 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3918 ms 100.0%
  mm 0.4056 ms 96.6%
  triton_mm_11523 0.4372 ms 89.6%
  triton_mm_11527 0.4467 ms 87.7%
  triton_mm_11524 0.4700 ms 83.4%
  triton_mm_11525 0.4796 ms 81.7%
  triton_mm_11531 0.4896 ms 80.0%
  triton_mm_11528 0.4908 ms 79.8%
  triton_mm_11520 0.5900 ms 66.4%
  triton_mm_11521 0.6126 ms 64.0%
SingleProcess AUTOTUNE benchmarking takes 2.1874 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8680 0.0317 ms 100.0%
  triton_mm_8676 0.0326 ms 97.3%
  triton_mm_8678 0.0327 ms 97.2%
  triton_mm_8675 0.0327 ms 97.1%
  triton_mm_8677 0.0327 ms 97.1%
  triton_mm_8681 0.0330 ms 96.1%
  triton_mm_8673 0.0344 ms 92.2%
  triton_mm_8671 0.0350 ms 90.7%
  triton_mm_8672 0.0352 ms 90.3%
  triton_mm_8684 0.0358 ms 88.6%
SingleProcess AUTOTUNE benchmarking takes 2.2442 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:33, 109.62s/it]AUTOTUNE mm(5148x2048, 2048x2048)
  triton_mm_11530 0.3895 ms 100.0%
  triton_mm_11523 0.4288 ms 90.8%
  triton_mm_11527 0.4407 ms 88.4%
  mm 0.4573 ms 85.2%
  triton_mm_11524 0.4630 ms 84.1%
  triton_mm_11525 0.4723 ms 82.5%
  triton_mm_11528 0.4853 ms 80.2%
  triton_mm_11531 0.4862 ms 80.1%
  triton_mm_11520 0.5912 ms 65.9%
  triton_mm_11521 0.6067 ms 64.2%
SingleProcess AUTOTUNE benchmarking takes 2.1510 seconds and 0.0000 seconds precompiling
AUTOTUNE mm(5130x28, 28x1920)
  triton_mm_8675 0.0317 ms 100.0%
  triton_mm_8676 0.0318 ms 99.7%
  triton_mm_8681 0.0324 ms 98.0%
  triton_mm_8677 0.0330 ms 96.1%
  triton_mm_8678 0.0330 ms 96.0%
  triton_mm_8680 0.0330 ms 96.0%
  triton_mm_8673 0.0349 ms 90.8%
  triton_mm_8672 0.0352 ms 90.2%
  triton_mm_8671 0.0353 ms 89.9%
  triton_mm_8684 0.0366 ms 86.7%
SingleProcess AUTOTUNE benchmarking takes 2.2087 seconds and 0.0000 seconds precompiling
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:34, 109.64s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   2%|2         | 1/44 [01:49<1:18:38, 109.73s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.32s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.32s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.32s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.32s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.32s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.32s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:07, 47.33s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   5%|4         | 2/44 [01:53<33:09, 47.37s/it]   /home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.32s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.32s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.32s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.32s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.32s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.32s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<17:59, 26.33s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:122: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   7%|6         | 3/44 [01:54<18:00, 26.35s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:55, 16.39s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:55, 16.39s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:55, 16.40s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:55, 16.39s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:55, 16.40s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:56, 16.40s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:55, 16.40s/it]
Validation Progress:   9%|9         | 4/44 [01:55<10:56, 16.42s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.88s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.88s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.88s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.88s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.88s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.89s/it]
Validation Progress:  11%|#1        | 5/44 [01:56<07:04, 10.89s/it]
Validation Progress:  11%|#1        | 5/44 [01:57<07:04, 10.89s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  14%|#3        | 6/44 [01:58<04:47,  7.57s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  16%|#5        | 7/44 [01:59<03:22,  5.48s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.11s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.11s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.11s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.11s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.11s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.12s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.11s/it]
Validation Progress:  18%|#8        | 8/44 [02:00<02:28,  4.12s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  20%|##        | 9/44 [02:01<01:52,  3.21s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:02<01:28,  2.60s/it]
Validation Progress:  23%|##2       | 10/44 [02:03<01:28,  2.60s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:11,  2.18s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:12,  2.19s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:10,  2.15s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:12,  2.19s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:12,  2.19s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:12,  2.18s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:11,  2.18s/it]
Validation Progress:  25%|##5       | 11/44 [02:04<01:12,  2.19s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<00:59,  1.87s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.89s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.90s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.90s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.90s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.89s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.89s/it]
Validation Progress:  27%|##7       | 12/44 [02:05<01:00,  1.90s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.69s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.70s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.70s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.71s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.71s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.70s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.71s/it]
Validation Progress:  30%|##9       | 13/44 [02:06<00:52,  1.71s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.55s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  32%|###1      | 14/44 [02:07<00:46,  1.56s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.45s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  34%|###4      | 15/44 [02:09<00:42,  1.46s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:38,  1.39s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.40s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.40s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.40s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.39s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.40s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.40s/it]
Validation Progress:  36%|###6      | 16/44 [02:10<00:39,  1.40s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  39%|###8      | 17/44 [02:11<00:37,  1.38s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.34s/it]
Validation Progress:  41%|####      | 18/44 [02:12<00:34,  1.35s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  43%|####3     | 19/44 [02:14<00:33,  1.34s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  45%|####5     | 20/44 [02:15<00:31,  1.31s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  48%|####7     | 21/44 [02:16<00:29,  1.28s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  50%|#####     | 22/44 [02:17<00:27,  1.26s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [02:19<00:26,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.30s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [02:20<00:26,  1.31s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  57%|#####6    | 25/44 [02:22<00:25,  1.33s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  59%|#####9    | 26/44 [02:23<00:23,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  61%|######1   | 27/44 [02:24<00:21,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  64%|######3   | 28/44 [02:25<00:20,  1.28s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.34s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.33s/it]
Validation Progress:  66%|######5   | 29/44 [02:27<00:20,  1.34s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [02:28<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  70%|#######   | 31/44 [02:29<00:17,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  73%|#######2  | 32/44 [02:31<00:16,  1.34s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.30s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.31s/it]
Validation Progress:  75%|#######5  | 33/44 [02:32<00:14,  1.31s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [02:33<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [02:35<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [02:36<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [02:37<00:09,  1.35s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.32s/it]
Validation Progress:  86%|########6 | 38/44 [02:39<00:07,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.33s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  89%|########8 | 39/44 [02:40<00:06,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [02:41<00:05,  1.34s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [02:43<00:03,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.30s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [02:44<00:02,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress:  98%|#########7| 43/44 [02:45<00:01,  1.32s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.18s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  1.19s/it]
Validation Progress: 100%|##########| 44/44 [02:46<00:00,  3.79s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:03<00:00,  3.37s/it]
100%|##########| 1/1 [00:03<00:00,  3.38s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|          | 1/157 [00:01<03:03,  1.17s/it]
  2%|1         | 3/157 [00:01<00:53,  2.86it/s]
  3%|3         | 5/157 [00:01<00:30,  5.05it/s]
  4%|4         | 7/157 [00:01<00:20,  7.27it/s]
  6%|5         | 9/157 [00:01<00:15,  9.39it/s]
  7%|7         | 11/157 [00:01<00:12, 11.29it/s]
  8%|8         | 13/157 [00:01<00:11, 12.63it/s]
 10%|9         | 15/157 [00:01<00:10, 13.99it/s]
 11%|#         | 17/157 [00:02<00:09, 15.04it/s]
 12%|#2        | 19/157 [00:02<00:08, 15.83it/s]
 13%|#3        | 21/157 [00:02<00:08, 15.86it/s]
 15%|#4        | 23/157 [00:02<00:08, 16.43it/s]
 16%|#5        | 25/157 [00:02<00:07, 16.87it/s]
 17%|#7        | 27/157 [00:02<00:07, 17.22it/s]
 18%|#8        | 29/157 [00:02<00:07, 17.33it/s]
 20%|#9        | 31/157 [00:02<00:07, 17.35it/s]
 21%|##1       | 33/157 [00:02<00:07, 17.48it/s]
 22%|##2       | 35/157 [00:03<00:06, 17.63it/s]
 24%|##3       | 37/157 [00:03<00:07, 16.53it/s]
 25%|##4       | 39/157 [00:03<00:07, 16.84it/s]
 26%|##6       | 41/157 [00:03<00:06, 16.96it/s]
 27%|##7       | 43/157 [00:03<00:06, 16.89it/s]
 29%|##8       | 45/157 [00:03<00:06, 16.53it/s]
 30%|##9       | 47/157 [00:03<00:06, 16.90it/s]
 31%|###1      | 49/157 [00:03<00:06, 17.00it/s]
 32%|###2      | 51/157 [00:04<00:06, 17.30it/s]
 34%|###3      | 53/157 [00:04<00:06, 16.86it/s]
 35%|###5      | 55/157 [00:04<00:05, 17.03it/s]
 36%|###6      | 57/157 [00:04<00:05, 17.21it/s]
 38%|###7      | 59/157 [00:04<00:05, 17.10it/s]
 39%|###8      | 61/157 [00:04<00:05, 16.79it/s]
 40%|####      | 63/157 [00:04<00:05, 17.06it/s]
 41%|####1     | 65/157 [00:04<00:05, 17.24it/s]
 43%|####2     | 67/157 [00:04<00:05, 17.09it/s]
 44%|####3     | 69/157 [00:05<00:05, 15.36it/s]
 45%|####5     | 71/157 [00:05<00:05, 16.07it/s]
 46%|####6     | 73/157 [00:05<00:05, 16.60it/s]
 48%|####7     | 75/157 [00:05<00:04, 16.96it/s]
 49%|####9     | 77/157 [00:05<00:04, 17.23it/s]
 50%|#####     | 79/157 [00:05<00:04, 17.41it/s]
 52%|#####1    | 81/157 [00:05<00:04, 17.55it/s]
 53%|#####2    | 83/157 [00:05<00:04, 17.69it/s]
 54%|#####4    | 85/157 [00:06<00:04, 17.35it/s]
 55%|#####5    | 87/157 [00:06<00:04, 17.29it/s]
 57%|#####6    | 89/157 [00:06<00:03, 17.11it/s]
 58%|#####7    | 91/157 [00:06<00:03, 16.78it/s]
 59%|#####9    | 93/157 [00:06<00:03, 16.41it/s]
 61%|######    | 95/157 [00:06<00:03, 16.61it/s]
 62%|######1   | 97/157 [00:06<00:03, 16.65it/s]
 63%|######3   | 99/157 [00:06<00:03, 16.95it/s]
 64%|######4   | 101/157 [00:07<00:03, 17.23it/s]
 66%|######5   | 103/157 [00:07<00:03, 17.45it/s]
 67%|######6   | 105/157 [00:07<00:02, 17.36it/s]
 68%|######8   | 107/157 [00:07<00:02, 17.01it/s]
 69%|######9   | 109/157 [00:07<00:02, 17.27it/s]
 71%|#######   | 111/157 [00:07<00:02, 17.42it/s]
 72%|#######1  | 113/157 [00:07<00:02, 17.11it/s]
 73%|#######3  | 115/157 [00:07<00:02, 17.32it/s]
 75%|#######4  | 117/157 [00:07<00:02, 17.43it/s]
 76%|#######5  | 119/157 [00:08<00:02, 17.26it/s]
 77%|#######7  | 121/157 [00:08<00:02, 16.89it/s]
 78%|#######8  | 123/157 [00:08<00:01, 17.16it/s]
 80%|#######9  | 125/157 [00:08<00:01, 17.36it/s]
 81%|########  | 127/157 [00:08<00:01, 17.44it/s]
 82%|########2 | 129/157 [00:08<00:01, 17.19it/s]
 83%|########3 | 131/157 [00:08<00:01, 17.41it/s]
 85%|########4 | 133/157 [00:08<00:01, 14.57it/s]
 86%|########5 | 135/157 [00:09<00:01, 15.40it/s]
 87%|########7 | 137/157 [00:09<00:01, 16.04it/s]
 89%|########8 | 139/157 [00:09<00:01, 15.89it/s]
 90%|########9 | 141/157 [00:09<00:00, 16.12it/s]
 91%|#########1| 143/157 [00:09<00:00, 16.25it/s]
 92%|#########2| 145/157 [00:09<00:00, 16.74it/s]
 94%|#########3| 147/157 [00:09<00:00, 17.06it/s]
 95%|#########4| 149/157 [00:09<00:00, 17.35it/s]
 96%|#########6| 151/157 [00:09<00:00, 17.57it/s]
 97%|#########7| 153/157 [00:10<00:00, 17.63it/s]
 99%|#########8| 155/157 [00:10<00:00, 17.45it/s]
100%|##########| 157/157 [00:11<00:00,  6.25it/s]
100%|##########| 157/157 [00:11<00:00, 14.27it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:45, 17.01it/s]
  1%|          | 4/781 [00:00<00:46, 16.62it/s]
  1%|          | 6/781 [00:00<00:47, 16.38it/s]
  1%|1         | 8/781 [00:00<00:46, 16.57it/s]
  1%|1         | 10/781 [00:00<00:46, 16.69it/s]
  2%|1         | 12/781 [00:00<00:46, 16.52it/s]
  2%|1         | 14/781 [00:00<00:47, 16.25it/s]
  2%|2         | 16/781 [00:00<00:46, 16.49it/s]
  2%|2         | 18/781 [00:01<00:45, 16.69it/s]
  3%|2         | 20/781 [00:01<00:45, 16.62it/s]
  3%|2         | 22/781 [00:01<00:45, 16.81it/s]
  3%|3         | 24/781 [00:01<00:44, 16.97it/s]
  3%|3         | 26/781 [00:01<00:44, 16.88it/s]
  4%|3         | 28/781 [00:01<00:45, 16.66it/s]
  4%|3         | 30/781 [00:01<00:46, 16.20it/s]
  4%|4         | 32/781 [00:01<00:46, 15.95it/s]
  4%|4         | 34/781 [00:02<00:47, 15.77it/s]
  5%|4         | 36/781 [00:02<00:47, 15.81it/s]
  5%|4         | 38/781 [00:02<00:47, 15.78it/s]
  5%|5         | 40/781 [00:02<00:46, 15.79it/s]
  5%|5         | 42/781 [00:02<00:46, 15.76it/s]
  6%|5         | 44/781 [00:02<00:46, 15.81it/s]
  6%|5         | 46/781 [00:02<00:46, 15.80it/s]
  6%|6         | 48/781 [00:02<00:46, 15.83it/s]
  6%|6         | 50/781 [00:03<00:46, 15.89it/s]
  7%|6         | 52/781 [00:03<00:46, 15.79it/s]
  7%|6         | 54/781 [00:03<00:45, 15.84it/s]
  7%|7         | 56/781 [00:03<00:45, 15.88it/s]
  7%|7         | 58/781 [00:03<00:45, 15.93it/s]
  8%|7         | 60/781 [00:03<00:45, 15.95it/s]
  8%|7         | 62/781 [00:03<00:45, 15.97it/s]
  8%|8         | 64/781 [00:03<00:44, 15.97it/s]
  8%|8         | 66/781 [00:04<00:44, 15.97it/s]
  9%|8         | 68/781 [00:04<00:44, 15.97it/s]
  9%|8         | 70/781 [00:04<00:44, 16.01it/s]
  9%|9         | 72/781 [00:04<00:44, 15.91it/s]
  9%|9         | 74/781 [00:04<00:44, 15.82it/s]
 10%|9         | 76/781 [00:04<00:44, 15.78it/s]
 10%|9         | 78/781 [00:04<00:44, 15.73it/s]
 10%|#         | 80/781 [00:04<00:44, 15.70it/s]
 10%|#         | 82/781 [00:05<00:44, 15.67it/s]
 11%|#         | 84/781 [00:05<00:44, 15.67it/s]
 11%|#1        | 86/781 [00:05<00:44, 15.63it/s]
 11%|#1        | 88/781 [00:05<00:44, 15.62it/s]
 12%|#1        | 90/781 [00:05<00:44, 15.65it/s]
 12%|#1        | 92/781 [00:05<00:44, 15.64it/s]
 12%|#2        | 94/781 [00:05<00:43, 15.64it/s]
 12%|#2        | 96/781 [00:05<00:43, 15.64it/s]
 13%|#2        | 98/781 [00:06<00:43, 15.62it/s]
 13%|#2        | 100/781 [00:06<00:43, 15.63it/s]
 13%|#3        | 102/781 [00:06<00:43, 15.67it/s]
 13%|#3        | 104/781 [00:06<00:43, 15.66it/s]
 14%|#3        | 106/781 [00:06<00:43, 15.66it/s]
 14%|#3        | 108/781 [00:06<00:43, 15.64it/s]
 14%|#4        | 110/781 [00:06<00:42, 15.65it/s]
 14%|#4        | 112/781 [00:07<00:42, 15.77it/s]
 15%|#4        | 114/781 [00:07<00:42, 15.83it/s]
 15%|#4        | 116/781 [00:07<00:41, 15.90it/s]
 15%|#5        | 118/781 [00:07<00:41, 15.85it/s]
 15%|#5        | 120/781 [00:07<00:41, 15.90it/s]
 16%|#5        | 122/781 [00:07<00:41, 15.93it/s]
 16%|#5        | 124/781 [00:07<00:41, 15.94it/s]
 16%|#6        | 126/781 [00:07<00:41, 15.86it/s]
 16%|#6        | 128/781 [00:08<00:41, 15.78it/s]
 17%|#6        | 130/781 [00:08<00:41, 15.69it/s]
 17%|#6        | 132/781 [00:08<00:41, 15.67it/s]
 17%|#7        | 134/781 [00:08<00:41, 15.65it/s]
 17%|#7        | 136/781 [00:08<00:41, 15.64it/s]
 18%|#7        | 138/781 [00:08<00:41, 15.65it/s]
 18%|#7        | 140/781 [00:08<00:41, 15.63it/s]
 18%|#8        | 142/781 [00:08<00:40, 15.63it/s]
 18%|#8        | 144/781 [00:09<00:40, 15.70it/s]
 19%|#8        | 146/781 [00:09<00:40, 15.73it/s]
 19%|#8        | 148/781 [00:09<00:40, 15.78it/s]
 19%|#9        | 150/781 [00:09<00:40, 15.68it/s]
 19%|#9        | 152/781 [00:09<00:39, 15.75it/s]
 20%|#9        | 154/781 [00:09<00:39, 15.82it/s]
 20%|#9        | 156/781 [00:09<00:39, 15.84it/s]
 20%|##        | 158/781 [00:09<00:39, 15.87it/s]
 20%|##        | 160/781 [00:10<00:39, 15.91it/s]
 21%|##        | 162/781 [00:10<00:38, 15.95it/s]
 21%|##        | 164/781 [00:10<00:38, 15.96it/s]
 21%|##1       | 166/781 [00:10<00:38, 15.97it/s]
 22%|##1       | 168/781 [00:10<00:38, 15.98it/s]
 22%|##1       | 170/781 [00:10<00:38, 16.00it/s]
 22%|##2       | 172/781 [00:10<00:38, 16.02it/s]
 22%|##2       | 174/781 [00:10<00:37, 16.02it/s]
 23%|##2       | 176/781 [00:11<00:37, 16.00it/s]
 23%|##2       | 178/781 [00:11<00:37, 16.01it/s]
 23%|##3       | 180/781 [00:11<00:37, 16.03it/s]
 23%|##3       | 182/781 [00:11<00:37, 16.00it/s]
 24%|##3       | 184/781 [00:11<00:37, 16.00it/s]
 24%|##3       | 186/781 [00:11<00:37, 16.00it/s]
 24%|##4       | 188/781 [00:11<00:37, 15.78it/s]
 24%|##4       | 190/781 [00:11<00:37, 15.66it/s]
 25%|##4       | 192/781 [00:12<00:37, 15.72it/s]
 25%|##4       | 194/781 [00:12<00:37, 15.71it/s]
 25%|##5       | 196/781 [00:12<00:37, 15.59it/s]
 25%|##5       | 198/781 [00:12<00:37, 15.54it/s]
 26%|##5       | 200/781 [00:12<00:37, 15.54it/s]
 26%|##5       | 202/781 [00:12<00:37, 15.59it/s]
 26%|##6       | 204/781 [00:12<00:37, 15.48it/s]
 26%|##6       | 206/781 [00:12<00:36, 15.67it/s]
 27%|##6       | 208/781 [00:13<00:36, 15.66it/s]
 27%|##6       | 210/781 [00:13<00:35, 15.92it/s]
 27%|##7       | 212/781 [00:13<00:35, 16.03it/s]
 27%|##7       | 214/781 [00:13<00:35, 15.90it/s]
 28%|##7       | 216/781 [00:13<00:35, 15.70it/s]
 28%|##7       | 218/781 [00:13<00:35, 15.88it/s]
 28%|##8       | 220/781 [00:13<00:35, 15.79it/s]
 28%|##8       | 222/781 [00:13<00:35, 15.94it/s]
 29%|##8       | 224/781 [00:14<00:34, 16.25it/s]
 29%|##8       | 226/781 [00:14<00:34, 16.31it/s]
 29%|##9       | 228/781 [00:14<00:34, 16.21it/s]
 29%|##9       | 230/781 [00:14<00:34, 16.15it/s]
 30%|##9       | 232/781 [00:14<00:34, 16.11it/s]
 30%|##9       | 234/781 [00:14<00:34, 16.08it/s]
 30%|###       | 236/781 [00:14<00:33, 16.21it/s]
 30%|###       | 238/781 [00:14<00:33, 16.38it/s]
 31%|###       | 240/781 [00:15<00:32, 16.74it/s]
 31%|###       | 242/781 [00:15<00:32, 16.50it/s]
 31%|###1      | 244/781 [00:15<00:32, 16.32it/s]
 31%|###1      | 246/781 [00:15<00:32, 16.35it/s]
 32%|###1      | 248/781 [00:15<00:32, 16.38it/s]
 32%|###2      | 250/781 [00:15<00:32, 16.43it/s]
 32%|###2      | 252/781 [00:15<00:32, 16.28it/s]
 33%|###2      | 254/781 [00:15<00:32, 16.20it/s]
 33%|###2      | 256/781 [00:16<00:31, 16.53it/s]
 33%|###3      | 258/781 [00:16<00:31, 16.62it/s]
 33%|###3      | 260/781 [00:16<00:30, 16.91it/s]
 34%|###3      | 262/781 [00:16<00:42, 12.18it/s]
 34%|###3      | 264/781 [00:16<00:39, 13.25it/s]
 34%|###4      | 266/781 [00:16<00:36, 13.98it/s]
 34%|###4      | 268/781 [00:16<00:35, 14.51it/s]
 35%|###4      | 270/781 [00:17<00:33, 15.32it/s]
 35%|###4      | 272/781 [00:17<00:31, 15.95it/s]
 35%|###5      | 274/781 [00:17<00:31, 16.18it/s]
 35%|###5      | 276/781 [00:17<00:31, 16.09it/s]
 36%|###5      | 278/781 [00:17<00:31, 16.05it/s]
 36%|###5      | 280/781 [00:17<00:30, 16.26it/s]
 36%|###6      | 282/781 [00:17<00:30, 16.39it/s]
 36%|###6      | 284/781 [00:17<00:30, 16.47it/s]
 37%|###6      | 286/781 [00:17<00:29, 16.50it/s]
 37%|###6      | 288/781 [00:18<00:29, 16.53it/s]
 37%|###7      | 290/781 [00:18<00:29, 16.55it/s]
 37%|###7      | 292/781 [00:18<00:29, 16.61it/s]
 38%|###7      | 294/781 [00:18<00:29, 16.44it/s]
 38%|###7      | 296/781 [00:18<00:29, 16.51it/s]
 38%|###8      | 298/781 [00:18<00:29, 16.37it/s]
 38%|###8      | 300/781 [00:18<00:29, 16.24it/s]
 39%|###8      | 302/781 [00:18<00:29, 16.33it/s]
 39%|###8      | 304/781 [00:19<00:29, 16.24it/s]
 39%|###9      | 306/781 [00:19<00:29, 16.18it/s]
 39%|###9      | 308/781 [00:19<00:29, 16.29it/s]
 40%|###9      | 310/781 [00:19<00:28, 16.64it/s]
 40%|###9      | 312/781 [00:19<00:28, 16.46it/s]
 40%|####      | 314/781 [00:19<00:28, 16.31it/s]
 40%|####      | 316/781 [00:19<00:28, 16.38it/s]
 41%|####      | 318/781 [00:19<00:28, 16.47it/s]
 41%|####      | 320/781 [00:20<00:28, 16.31it/s]
 41%|####1     | 322/781 [00:20<00:28, 16.27it/s]
 41%|####1     | 324/781 [00:20<00:27, 16.41it/s]
 42%|####1     | 326/781 [00:20<00:27, 16.27it/s]
 42%|####1     | 328/781 [00:20<00:27, 16.63it/s]
 42%|####2     | 330/781 [00:20<00:27, 16.42it/s]
 43%|####2     | 332/781 [00:20<00:27, 16.53it/s]
 43%|####2     | 334/781 [00:20<00:26, 16.58it/s]
 43%|####3     | 336/781 [00:21<00:26, 16.53it/s]
 43%|####3     | 338/781 [00:21<00:27, 16.22it/s]
 44%|####3     | 340/781 [00:21<00:27, 16.25it/s]
 44%|####3     | 342/781 [00:21<00:27, 16.06it/s]
 44%|####4     | 344/781 [00:21<00:27, 15.92it/s]
 44%|####4     | 346/781 [00:21<00:27, 16.05it/s]
 45%|####4     | 348/781 [00:21<00:27, 15.90it/s]
 45%|####4     | 350/781 [00:21<00:27, 15.95it/s]
 45%|####5     | 352/781 [00:22<00:27, 15.78it/s]
 45%|####5     | 354/781 [00:22<00:27, 15.81it/s]
 46%|####5     | 356/781 [00:22<00:26, 15.79it/s]
 46%|####5     | 358/781 [00:22<00:26, 15.70it/s]
 46%|####6     | 360/781 [00:22<00:26, 15.76it/s]
 46%|####6     | 362/781 [00:22<00:26, 15.64it/s]
 47%|####6     | 364/781 [00:22<00:26, 15.64it/s]
 47%|####6     | 366/781 [00:22<00:26, 15.69it/s]
 47%|####7     | 368/781 [00:23<00:25, 15.95it/s]
 47%|####7     | 370/781 [00:23<00:25, 15.91it/s]
 48%|####7     | 372/781 [00:23<00:25, 15.99it/s]
 48%|####7     | 374/781 [00:23<00:24, 16.40it/s]
 48%|####8     | 376/781 [00:23<00:24, 16.50it/s]
 48%|####8     | 378/781 [00:23<00:24, 16.53it/s]
 49%|####8     | 380/781 [00:23<00:24, 16.58it/s]
 49%|####8     | 382/781 [00:23<00:24, 16.37it/s]
 49%|####9     | 384/781 [00:24<00:24, 16.23it/s]
 49%|####9     | 386/781 [00:24<00:24, 16.13it/s]
 50%|####9     | 388/781 [00:24<00:24, 16.07it/s]
 50%|####9     | 390/781 [00:24<00:24, 16.24it/s]
 50%|#####     | 392/781 [00:24<00:24, 16.12it/s]
 50%|#####     | 394/781 [00:24<00:24, 16.08it/s]
 51%|#####     | 396/781 [00:24<00:23, 16.05it/s]
 51%|#####     | 398/781 [00:24<00:23, 16.48it/s]
 51%|#####1    | 400/781 [00:25<00:23, 16.34it/s]
 51%|#####1    | 402/781 [00:25<00:22, 16.69it/s]
 52%|#####1    | 404/781 [00:25<00:22, 16.49it/s]
 52%|#####1    | 406/781 [00:25<00:22, 16.81it/s]
 52%|#####2    | 408/781 [00:25<00:22, 16.79it/s]
 52%|#####2    | 410/781 [00:25<00:22, 16.81it/s]
 53%|#####2    | 412/781 [00:25<00:21, 16.78it/s]
 53%|#####3    | 414/781 [00:25<00:21, 16.74it/s]
 53%|#####3    | 416/781 [00:25<00:21, 16.75it/s]
 54%|#####3    | 418/781 [00:26<00:21, 16.53it/s]
 54%|#####3    | 420/781 [00:26<00:22, 16.38it/s]
 54%|#####4    | 422/781 [00:26<00:22, 16.26it/s]
 54%|#####4    | 424/781 [00:26<00:22, 16.17it/s]
 55%|#####4    | 426/781 [00:26<00:21, 16.32it/s]
 55%|#####4    | 428/781 [00:26<00:21, 16.20it/s]
 55%|#####5    | 430/781 [00:26<00:21, 16.33it/s]
 55%|#####5    | 432/781 [00:26<00:21, 16.44it/s]
 56%|#####5    | 434/781 [00:27<00:21, 16.29it/s]
 56%|#####5    | 436/781 [00:27<00:21, 16.23it/s]
 56%|#####6    | 438/781 [00:27<00:20, 16.63it/s]
 56%|#####6    | 440/781 [00:27<00:20, 16.65it/s]
 57%|#####6    | 442/781 [00:27<00:20, 16.44it/s]
 57%|#####6    | 444/781 [00:27<00:20, 16.80it/s]
 57%|#####7    | 446/781 [00:27<00:19, 16.77it/s]
 57%|#####7    | 448/781 [00:27<00:19, 16.65it/s]
 58%|#####7    | 450/781 [00:28<00:20, 16.39it/s]
 58%|#####7    | 452/781 [00:28<00:19, 16.57it/s]
 58%|#####8    | 454/781 [00:28<00:19, 16.72it/s]
 58%|#####8    | 456/781 [00:28<00:19, 16.45it/s]
 59%|#####8    | 458/781 [00:28<00:19, 16.74it/s]
 59%|#####8    | 460/781 [00:28<00:19, 16.63it/s]
 59%|#####9    | 462/781 [00:28<00:19, 16.33it/s]
 59%|#####9    | 464/781 [00:28<00:19, 16.38it/s]
 60%|#####9    | 466/781 [00:29<00:19, 16.34it/s]
 60%|#####9    | 468/781 [00:29<00:19, 16.36it/s]
 60%|######    | 470/781 [00:29<00:19, 16.13it/s]
 60%|######    | 472/781 [00:29<00:19, 16.02it/s]
 61%|######    | 474/781 [00:29<00:19, 15.91it/s]
 61%|######    | 476/781 [00:29<00:19, 15.82it/s]
 61%|######1   | 478/781 [00:29<00:18, 16.31it/s]
 61%|######1   | 480/781 [00:29<00:18, 16.43it/s]
 62%|######1   | 482/781 [00:30<00:18, 16.53it/s]
 62%|######1   | 484/781 [00:30<00:17, 16.62it/s]
 62%|######2   | 486/781 [00:30<00:17, 16.91it/s]
 62%|######2   | 488/781 [00:30<00:17, 16.87it/s]
 63%|######2   | 490/781 [00:30<00:17, 16.86it/s]
 63%|######2   | 492/781 [00:30<00:17, 16.85it/s]
 63%|######3   | 494/781 [00:30<00:17, 16.81it/s]
 64%|######3   | 496/781 [00:30<00:16, 17.04it/s]
 64%|######3   | 498/781 [00:30<00:16, 17.20it/s]
 64%|######4   | 500/781 [00:31<00:16, 17.10it/s]
 64%|######4   | 502/781 [00:31<00:16, 17.26it/s]
 65%|######4   | 504/781 [00:31<00:16, 17.12it/s]
 65%|######4   | 506/781 [00:31<00:16, 16.74it/s]
 65%|######5   | 508/781 [00:31<00:16, 16.74it/s]
 65%|######5   | 510/781 [00:31<00:15, 16.99it/s]
 66%|######5   | 512/781 [00:31<00:15, 17.05it/s]
 66%|######5   | 514/781 [00:31<00:15, 16.85it/s]
 66%|######6   | 516/781 [00:32<00:16, 16.53it/s]
 66%|######6   | 518/781 [00:32<00:16, 16.42it/s]
 67%|######6   | 520/781 [00:32<00:26,  9.76it/s]
 67%|######6   | 522/781 [00:32<00:23, 11.06it/s]
 67%|######7   | 524/781 [00:32<00:21, 12.06it/s]
 67%|######7   | 526/781 [00:32<00:19, 13.20it/s]
 68%|######7   | 528/781 [00:33<00:17, 14.13it/s]
 68%|######7   | 530/781 [00:33<00:16, 14.91it/s]
 68%|######8   | 532/781 [00:33<00:16, 15.56it/s]
 68%|######8   | 534/781 [00:33<00:15, 15.86it/s]
 69%|######8   | 536/781 [00:33<00:15, 16.08it/s]
 69%|######8   | 538/781 [00:33<00:14, 16.49it/s]
 69%|######9   | 540/781 [00:33<00:14, 16.74it/s]
 69%|######9   | 542/781 [00:33<00:14, 16.73it/s]
 70%|######9   | 544/781 [00:33<00:14, 16.65it/s]
 70%|######9   | 546/781 [00:34<00:13, 16.87it/s]
 70%|#######   | 548/781 [00:34<00:14, 16.61it/s]
 70%|#######   | 550/781 [00:34<00:13, 16.61it/s]
 71%|#######   | 552/781 [00:34<00:13, 16.76it/s]
 71%|#######   | 554/781 [00:34<00:13, 16.87it/s]
 71%|#######1  | 556/781 [00:34<00:13, 16.82it/s]
 71%|#######1  | 558/781 [00:34<00:13, 16.79it/s]
 72%|#######1  | 560/781 [00:34<00:13, 16.77it/s]
 72%|#######1  | 562/781 [00:35<00:12, 17.01it/s]
 72%|#######2  | 564/781 [00:35<00:12, 16.88it/s]
 72%|#######2  | 566/781 [00:35<00:12, 17.07it/s]
 73%|#######2  | 568/781 [00:35<00:12, 17.22it/s]
 73%|#######2  | 570/781 [00:35<00:12, 17.32it/s]
 73%|#######3  | 572/781 [00:35<00:12, 17.13it/s]
 73%|#######3  | 574/781 [00:35<00:12, 16.86it/s]
 74%|#######3  | 576/781 [00:35<00:12, 16.94it/s]
 74%|#######4  | 578/781 [00:35<00:12, 16.52it/s]
 74%|#######4  | 580/781 [00:36<00:12, 16.70it/s]
 75%|#######4  | 582/781 [00:36<00:11, 16.91it/s]
 75%|#######4  | 584/781 [00:36<00:11, 16.86it/s]
 75%|#######5  | 586/781 [00:36<00:11, 16.83it/s]
 75%|#######5  | 588/781 [00:36<00:11, 16.80it/s]
 76%|#######5  | 590/781 [00:36<00:11, 17.02it/s]
 76%|#######5  | 592/781 [00:36<00:11, 17.17it/s]
 76%|#######6  | 594/781 [00:36<00:10, 17.15it/s]
 76%|#######6  | 596/781 [00:37<00:10, 17.03it/s]
 77%|#######6  | 598/781 [00:37<00:10, 17.22it/s]
 77%|#######6  | 600/781 [00:37<00:10, 16.82it/s]
 77%|#######7  | 602/781 [00:37<00:10, 17.06it/s]
 77%|#######7  | 604/781 [00:37<00:10, 17.01it/s]
 78%|#######7  | 606/781 [00:37<00:10, 16.70it/s]
 78%|#######7  | 608/781 [00:37<00:10, 16.73it/s]
 78%|#######8  | 610/781 [00:37<00:10, 16.43it/s]
 78%|#######8  | 612/781 [00:38<00:10, 16.42it/s]
 79%|#######8  | 614/781 [00:38<00:10, 16.26it/s]
 79%|#######8  | 616/781 [00:38<00:09, 16.58it/s]
 79%|#######9  | 618/781 [00:38<00:09, 16.40it/s]
 79%|#######9  | 620/781 [00:38<00:09, 16.47it/s]
 80%|#######9  | 622/781 [00:38<00:09, 16.27it/s]
 80%|#######9  | 624/781 [00:38<00:09, 16.51it/s]
 80%|########  | 626/781 [00:38<00:09, 16.24it/s]
 80%|########  | 628/781 [00:38<00:09, 16.45it/s]
 81%|########  | 630/781 [00:39<00:09, 16.20it/s]
 81%|########  | 632/781 [00:39<00:09, 16.27it/s]
 81%|########1 | 634/781 [00:39<00:08, 16.48it/s]
 81%|########1 | 636/781 [00:39<00:08, 16.17it/s]
 82%|########1 | 638/781 [00:39<00:08, 16.40it/s]
 82%|########1 | 640/781 [00:39<00:08, 16.67it/s]
 82%|########2 | 642/781 [00:39<00:08, 16.66it/s]
 82%|########2 | 644/781 [00:39<00:08, 16.69it/s]
 83%|########2 | 646/781 [00:40<00:08, 16.87it/s]
 83%|########2 | 648/781 [00:40<00:07, 17.04it/s]
 83%|########3 | 650/781 [00:40<00:07, 16.96it/s]
 83%|########3 | 652/781 [00:40<00:07, 17.14it/s]
 84%|########3 | 654/781 [00:40<00:07, 16.74it/s]
 84%|########3 | 656/781 [00:40<00:07, 16.72it/s]
 84%|########4 | 658/781 [00:40<00:07, 16.92it/s]
 85%|########4 | 660/781 [00:40<00:07, 16.86it/s]
 85%|########4 | 662/781 [00:41<00:07, 16.81it/s]
 85%|########5 | 664/781 [00:41<00:06, 17.02it/s]
 85%|########5 | 666/781 [00:41<00:06, 17.12it/s]
 86%|########5 | 668/781 [00:41<00:06, 17.19it/s]
 86%|########5 | 670/781 [00:41<00:06, 17.24it/s]
 86%|########6 | 672/781 [00:41<00:06, 16.95it/s]
 86%|########6 | 674/781 [00:41<00:06, 16.94it/s]
 87%|########6 | 676/781 [00:41<00:06, 16.82it/s]
 87%|########6 | 678/781 [00:41<00:06, 16.85it/s]
 87%|########7 | 680/781 [00:42<00:06, 16.30it/s]
 87%|########7 | 682/781 [00:42<00:06, 16.26it/s]
 88%|########7 | 684/781 [00:42<00:05, 16.32it/s]
 88%|########7 | 686/781 [00:42<00:05, 16.36it/s]
 88%|########8 | 688/781 [00:42<00:05, 16.29it/s]
 88%|########8 | 690/781 [00:42<00:05, 16.33it/s]
 89%|########8 | 692/781 [00:42<00:05, 16.06it/s]
 89%|########8 | 694/781 [00:42<00:05, 16.20it/s]
 89%|########9 | 696/781 [00:43<00:05, 16.19it/s]
 89%|########9 | 698/781 [00:43<00:05, 16.21it/s]
 90%|########9 | 700/781 [00:43<00:04, 16.56it/s]
 90%|########9 | 702/781 [00:43<00:04, 16.40it/s]
 90%|######### | 704/781 [00:43<00:04, 16.18it/s]
 90%|######### | 706/781 [00:43<00:04, 16.00it/s]
 91%|######### | 708/781 [00:43<00:04, 15.93it/s]
 91%|######### | 710/781 [00:43<00:04, 15.92it/s]
 91%|#########1| 712/781 [00:44<00:04, 15.92it/s]
 91%|#########1| 714/781 [00:44<00:04, 16.10it/s]
 92%|#########1| 716/781 [00:44<00:04, 16.03it/s]
 92%|#########1| 718/781 [00:44<00:03, 16.24it/s]
 92%|#########2| 720/781 [00:44<00:03, 16.15it/s]
 92%|#########2| 722/781 [00:44<00:03, 16.29it/s]
 93%|#########2| 724/781 [00:44<00:03, 16.42it/s]
 93%|#########2| 726/781 [00:44<00:03, 16.27it/s]
 93%|#########3| 728/781 [00:45<00:03, 16.15it/s]
 93%|#########3| 730/781 [00:45<00:03, 16.25it/s]
 94%|#########3| 732/781 [00:45<00:02, 16.60it/s]
 94%|#########3| 734/781 [00:45<00:02, 16.87it/s]
 94%|#########4| 736/781 [00:45<00:02, 16.57it/s]
 94%|#########4| 738/781 [00:45<00:02, 16.35it/s]
 95%|#########4| 740/781 [00:45<00:02, 16.13it/s]
 95%|#########5| 742/781 [00:45<00:02, 15.97it/s]
 95%|#########5| 744/781 [00:46<00:02, 15.88it/s]
 96%|#########5| 746/781 [00:46<00:02, 16.04it/s]
 96%|#########5| 748/781 [00:46<00:02, 16.02it/s]
 96%|#########6| 750/781 [00:46<00:01, 15.97it/s]
 96%|#########6| 752/781 [00:46<00:01, 15.96it/s]
 97%|#########6| 754/781 [00:46<00:01, 16.18it/s]
 97%|#########6| 756/781 [00:46<00:01, 15.99it/s]
 97%|#########7| 758/781 [00:46<00:01, 16.08it/s]
 97%|#########7| 760/781 [00:47<00:01, 16.38it/s]
 98%|#########7| 762/781 [00:47<00:01, 16.34it/s]
 98%|#########7| 764/781 [00:47<00:01, 16.13it/s]
 98%|#########8| 766/781 [00:47<00:00, 16.01it/s]
 98%|#########8| 768/781 [00:47<00:00, 16.17it/s]
 99%|#########8| 770/781 [00:47<00:00, 16.58it/s]
 99%|#########8| 772/781 [00:47<00:00, 16.61it/s]
 99%|#########9| 774/781 [00:47<00:00, 16.39it/s]
 99%|#########9| 776/781 [00:48<00:00, 16.24it/s]
100%|#########9| 778/781 [00:48<00:00, 16.59it/s]
100%|#########9| 780/781 [00:48<00:00, 16.38it/s]
100%|##########| 781/781 [00:48<00:00, 16.16it/s]
/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

/home/user/VAR/trainer.py:170: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   0%|          | 0/44 [00:00<?, ?it/s]
Validation Progress:   2%|2         | 1/44 [00:05<03:46,  5.26s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:46,  5.27s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:46,  5.26s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:46,  5.27s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:46,  5.28s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:46,  5.28s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:47,  5.29s/it]
Validation Progress:   2%|2         | 1/44 [00:05<03:51,  5.38s/it]
Validation Progress:   5%|4         | 2/44 [00:06<01:59,  2.86s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.86s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.86s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.86s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.87s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.87s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.87s/it]
Validation Progress:   5%|4         | 2/44 [00:06<02:00,  2.88s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:28,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:28,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:28,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:28,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:28,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:29,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:29,  2.17s/it]
Validation Progress:   7%|6         | 3/44 [00:07<01:30,  2.20s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:10,  1.77s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:10,  1.77s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:10,  1.77s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:10,  1.77s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:10,  1.77s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:10,  1.77s/it]
Validation Progress:   9%|9         | 4/44 [00:08<01:11,  1.78s/it]
Validation Progress:   9%|9         | 4/44 [00:09<01:11,  1.79s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.55s/it]
Validation Progress:  11%|#1        | 5/44 [00:10<01:00,  1.56s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.41s/it]
Validation Progress:  14%|#3        | 6/44 [00:11<00:53,  1.42s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:49,  1.35s/it]
Validation Progress:  16%|#5        | 7/44 [00:12<00:50,  1.35s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.29s/it]
Validation Progress:  18%|#8        | 8/44 [00:13<00:46,  1.30s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.27s/it]
Validation Progress:  20%|##        | 9/44 [00:14<00:44,  1.28s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:43,  1.27s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:42,  1.26s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:44,  1.30s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:43,  1.27s/it]
Validation Progress:  23%|##2       | 10/44 [00:16<00:44,  1.30s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:41,  1.27s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:41,  1.27s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:42,  1.28s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:42,  1.28s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:42,  1.28s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:41,  1.25s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:42,  1.28s/it]
Validation Progress:  25%|##5       | 11/44 [00:17<00:42,  1.29s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.28s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.28s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.28s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.27s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.25s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.27s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:40,  1.27s/it]
Validation Progress:  27%|##7       | 12/44 [00:18<00:41,  1.28s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.26s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.26s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:38,  1.25s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.27s/it]
Validation Progress:  30%|##9       | 13/44 [00:19<00:39,  1.27s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.25s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.26s/it]
Validation Progress:  32%|###1      | 14/44 [00:21<00:37,  1.26s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:36,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:35,  1.23s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:35,  1.24s/it]
Validation Progress:  34%|###4      | 15/44 [00:22<00:36,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  36%|###6      | 16/44 [00:23<00:34,  1.24s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.27s/it]
Validation Progress:  39%|###8      | 17/44 [00:24<00:34,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  41%|####      | 18/44 [00:26<00:32,  1.26s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:31,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:31,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:31,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:31,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:32,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:31,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:31,  1.28s/it]
Validation Progress:  43%|####3     | 19/44 [00:27<00:32,  1.28s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  45%|####5     | 20/44 [00:28<00:30,  1.26s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  48%|####7     | 21/44 [00:29<00:28,  1.24s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  50%|#####     | 22/44 [00:31<00:27,  1.23s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.29s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.28s/it]
Validation Progress:  52%|#####2    | 23/44 [00:32<00:26,  1.28s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  55%|#####4    | 24/44 [00:33<00:26,  1.31s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.34s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.35s/it]
Validation Progress:  57%|#####6    | 25/44 [00:35<00:25,  1.35s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.31s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.31s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.31s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.30s/it]
Validation Progress:  59%|#####9    | 26/44 [00:36<00:23,  1.30s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  61%|######1   | 27/44 [00:37<00:21,  1.29s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  64%|######3   | 28/44 [00:39<00:20,  1.27s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  66%|######5   | 29/44 [00:40<00:19,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.32s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.33s/it]
Validation Progress:  68%|######8   | 30/44 [00:41<00:18,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.32s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  70%|#######   | 31/44 [00:43<00:17,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  73%|#######2  | 32/44 [00:44<00:15,  1.33s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.29s/it]
Validation Progress:  75%|#######5  | 33/44 [00:45<00:14,  1.30s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  77%|#######7  | 34/44 [00:46<00:12,  1.28s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  80%|#######9  | 35/44 [00:48<00:11,  1.30s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  82%|########1 | 36/44 [00:49<00:10,  1.31s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.35s/it]
Validation Progress:  84%|########4 | 37/44 [00:51<00:09,  1.36s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.31s/it]
Validation Progress:  86%|########6 | 38/44 [00:52<00:07,  1.32s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  89%|########8 | 39/44 [00:53<00:06,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.34s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  91%|######### | 40/44 [00:55<00:05,  1.35s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  93%|#########3| 41/44 [00:56<00:03,  1.33s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  95%|#########5| 42/44 [00:57<00:02,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]
Validation Progress:  98%|#########7| 43/44 [00:58<00:01,  1.31s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.90s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]/home/user/VAR/tokenizer/tokenizer_image/dino_enc/dinov2.py:273: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(enabled=False):

Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.91s/it]
Validation Progress: 100%|##########| 44/44 [01:02<00:00,  1.41s/it]

  0%|          | 0/1 [00:00<?, ?it/s]
100%|##########| 1/1 [00:02<00:00,  2.44s/it]
100%|##########| 1/1 [00:02<00:00,  2.44s/it]

  0%|          | 0/157 [00:00<?, ?it/s]
  1%|1         | 2/157 [00:00<00:11, 13.83it/s]
  3%|2         | 4/157 [00:00<00:09, 15.37it/s]
  4%|3         | 6/157 [00:00<00:09, 15.83it/s]
  5%|5         | 8/157 [00:00<00:09, 16.20it/s]
  6%|6         | 10/157 [00:00<00:08, 16.45it/s]
  8%|7         | 12/157 [00:00<00:08, 16.57it/s]
  9%|8         | 14/157 [00:00<00:08, 16.41it/s]
 10%|#         | 16/157 [00:00<00:08, 16.56it/s]
 11%|#1        | 18/157 [00:01<00:08, 16.45it/s]
 13%|#2        | 20/157 [00:01<00:08, 16.60it/s]
 14%|#4        | 22/157 [00:01<00:08, 16.48it/s]
 15%|#5        | 24/157 [00:01<00:08, 16.49it/s]
 17%|#6        | 26/157 [00:01<00:07, 16.48it/s]
 18%|#7        | 28/157 [00:01<00:07, 16.49it/s]
 19%|#9        | 30/157 [00:01<00:07, 16.51it/s]
 20%|##        | 32/157 [00:01<00:07, 16.49it/s]
 22%|##1       | 34/157 [00:02<00:07, 16.51it/s]
 23%|##2       | 36/157 [00:02<00:07, 16.62it/s]
 24%|##4       | 38/157 [00:02<00:07, 16.60it/s]
 25%|##5       | 40/157 [00:02<00:07, 16.28it/s]
 27%|##6       | 42/157 [00:02<00:06, 16.55it/s]
 28%|##8       | 44/157 [00:02<00:06, 16.56it/s]
 29%|##9       | 46/157 [00:02<00:06, 16.78it/s]
 31%|###       | 48/157 [00:02<00:06, 16.72it/s]
 32%|###1      | 50/157 [00:03<00:06, 16.91it/s]
 33%|###3      | 52/157 [00:03<00:06, 16.80it/s]
 34%|###4      | 54/157 [00:03<00:06, 16.98it/s]
 36%|###5      | 56/157 [00:03<00:05, 16.90it/s]
 37%|###6      | 58/157 [00:03<00:05, 16.94it/s]
 38%|###8      | 60/157 [00:03<00:05, 16.79it/s]
 39%|###9      | 62/157 [00:03<00:05, 16.61it/s]
 41%|####      | 64/157 [00:03<00:05, 16.60it/s]
 42%|####2     | 66/157 [00:03<00:05, 16.64it/s]
 43%|####3     | 68/157 [00:04<00:05, 16.61it/s]
 45%|####4     | 70/157 [00:04<00:05, 16.59it/s]
 46%|####5     | 72/157 [00:04<00:05, 16.58it/s]
 47%|####7     | 74/157 [00:04<00:05, 16.60it/s]
 48%|####8     | 76/157 [00:04<00:04, 16.75it/s]
 50%|####9     | 78/157 [00:04<00:04, 16.48it/s]
 51%|#####     | 80/157 [00:04<00:04, 16.57it/s]
 52%|#####2    | 82/157 [00:04<00:04, 16.34it/s]
 54%|#####3    | 84/157 [00:05<00:05, 13.44it/s]
 55%|#####4    | 86/157 [00:05<00:04, 14.35it/s]
 56%|#####6    | 88/157 [00:05<00:04, 14.88it/s]
 57%|#####7    | 90/157 [00:05<00:04, 15.42it/s]
 59%|#####8    | 92/157 [00:05<00:04, 15.87it/s]
 60%|#####9    | 94/157 [00:05<00:03, 16.05it/s]
 61%|######1   | 96/157 [00:05<00:03, 16.40it/s]
 62%|######2   | 98/157 [00:06<00:03, 16.43it/s]
 64%|######3   | 100/157 [00:06<00:03, 16.69it/s]
 65%|######4   | 102/157 [00:06<00:03, 16.65it/s]
 66%|######6   | 104/157 [00:06<00:03, 16.76it/s]
 68%|######7   | 106/157 [00:06<00:03, 16.98it/s]
 69%|######8   | 108/157 [00:06<00:02, 17.06it/s]
 70%|#######   | 110/157 [00:06<00:02, 16.90it/s]
 71%|#######1  | 112/157 [00:06<00:02, 16.87it/s]
 73%|#######2  | 114/157 [00:06<00:02, 17.00it/s]
 74%|#######3  | 116/157 [00:07<00:02, 16.83it/s]
 75%|#######5  | 118/157 [00:07<00:02, 16.99it/s]
 76%|#######6  | 120/157 [00:07<00:02, 16.96it/s]
 78%|#######7  | 122/157 [00:07<00:02, 16.96it/s]
 79%|#######8  | 124/157 [00:07<00:01, 16.82it/s]
 80%|########  | 126/157 [00:07<00:01, 16.82it/s]
 82%|########1 | 128/157 [00:07<00:01, 16.84it/s]
 83%|########2 | 130/157 [00:07<00:01, 16.98it/s]
 84%|########4 | 132/157 [00:08<00:01, 17.00it/s]
 85%|########5 | 134/157 [00:08<00:01, 16.83it/s]
 87%|########6 | 136/157 [00:08<00:01, 16.87it/s]
 88%|########7 | 138/157 [00:08<00:01, 16.78it/s]
 89%|########9 | 140/157 [00:08<00:01, 16.70it/s]
 90%|######### | 142/157 [00:08<00:00, 16.69it/s]
 92%|#########1| 144/157 [00:08<00:00, 16.78it/s]
 93%|#########2| 146/157 [00:08<00:00, 16.73it/s]
 94%|#########4| 148/157 [00:08<00:00, 16.70it/s]
 96%|#########5| 150/157 [00:09<00:00, 16.56it/s]
 97%|#########6| 152/157 [00:09<00:00, 16.71it/s]
 98%|#########8| 154/157 [00:09<00:00, 16.72it/s]
 99%|#########9| 156/157 [00:09<00:00, 16.57it/s]
100%|##########| 157/157 [00:09<00:00, 16.56it/s]

  0%|          | 0/781 [00:00<?, ?it/s]
  0%|          | 2/781 [00:00<00:58, 13.25it/s]
  1%|          | 4/781 [00:00<00:49, 15.79it/s]
  1%|          | 6/781 [00:00<00:45, 17.01it/s]
  1%|1         | 8/781 [00:00<00:43, 17.62it/s]
  1%|1         | 10/781 [00:00<00:43, 17.74it/s]
  2%|1         | 12/781 [00:00<00:42, 17.95it/s]
  2%|1         | 14/781 [00:00<00:42, 18.22it/s]
  2%|2         | 16/781 [00:00<00:42, 18.09it/s]
  2%|2         | 18/781 [00:01<00:41, 18.31it/s]
  3%|2         | 20/781 [00:01<00:41, 18.14it/s]
  3%|2         | 22/781 [00:01<00:41, 18.21it/s]
  3%|3         | 24/781 [00:01<00:41, 18.30it/s]
  3%|3         | 26/781 [00:01<00:41, 18.29it/s]
  4%|3         | 28/781 [00:01<00:40, 18.46it/s]
  4%|3         | 30/781 [00:01<00:40, 18.45it/s]
  4%|4         | 32/781 [00:01<00:40, 18.29it/s]
  4%|4         | 34/781 [00:01<00:40, 18.28it/s]
  5%|4         | 36/781 [00:02<00:40, 18.46it/s]
  5%|4         | 38/781 [00:02<00:40, 18.37it/s]
  5%|5         | 40/781 [00:02<00:40, 18.39it/s]
  5%|5         | 42/781 [00:02<00:40, 18.32it/s]
  6%|5         | 44/781 [00:02<00:39, 18.49it/s]
  6%|5         | 46/781 [00:02<00:39, 18.49it/s]
  6%|6         | 48/781 [00:02<00:39, 18.42it/s]
  6%|6         | 50/781 [00:02<00:39, 18.54it/s]
  7%|6         | 52/781 [00:02<00:39, 18.49it/s]
  7%|6         | 54/781 [00:02<00:39, 18.28it/s]
  7%|7         | 56/781 [00:03<00:39, 18.39it/s]
  7%|7         | 58/781 [00:03<00:39, 18.34it/s]
  8%|7         | 60/781 [00:03<00:39, 18.33it/s]
  8%|7         | 62/781 [00:03<00:38, 18.53it/s]
  8%|8         | 64/781 [00:03<00:39, 18.34it/s]
  8%|8         | 66/781 [00:03<00:38, 18.48it/s]
  9%|8         | 68/781 [00:03<00:38, 18.46it/s]
  9%|8         | 70/781 [00:03<00:38, 18.47it/s]
  9%|9         | 72/781 [00:03<00:38, 18.47it/s]
  9%|9         | 74/781 [00:04<00:38, 18.33it/s]
 10%|9         | 76/781 [00:04<00:37, 18.56it/s]
 10%|9         | 78/781 [00:04<00:38, 18.36it/s]
 10%|#         | 80/781 [00:04<00:37, 18.53it/s]
 10%|#         | 82/781 [00:04<00:37, 18.51it/s]
 11%|#         | 84/781 [00:04<00:37, 18.56it/s]
 11%|#1        | 86/781 [00:04<00:37, 18.50it/s]
 11%|#1        | 88/781 [00:04<00:37, 18.51it/s]
 12%|#1        | 90/781 [00:04<00:37, 18.45it/s]
 12%|#1        | 92/781 [00:05<00:37, 18.43it/s]
 12%|#2        | 94/781 [00:05<00:37, 18.35it/s]
 12%|#2        | 96/781 [00:05<00:37, 18.50it/s]
 13%|#2        | 98/781 [00:05<00:36, 18.64it/s]
 13%|#2        | 100/781 [00:05<00:36, 18.45it/s]
 13%|#3        | 102/781 [00:05<00:36, 18.54it/s]
 13%|#3        | 104/781 [00:05<00:36, 18.57it/s]
 14%|#3        | 106/781 [00:05<00:36, 18.58it/s]
 14%|#3        | 108/781 [00:05<00:36, 18.48it/s]
 14%|#4        | 110/781 [00:06<00:36, 18.37it/s]
 14%|#4        | 112/781 [00:06<00:36, 18.24it/s]
 15%|#4        | 114/781 [00:06<00:36, 18.39it/s]
 15%|#4        | 116/781 [00:06<00:36, 18.13it/s]
 15%|#5        | 118/781 [00:06<00:36, 18.19it/s]
 15%|#5        | 120/781 [00:06<00:36, 18.25it/s]
 16%|#5        | 122/781 [00:06<00:35, 18.31it/s]
 16%|#5        | 124/781 [00:06<00:35, 18.35it/s]
 16%|#6        | 126/781 [00:06<00:35, 18.51it/s]
 16%|#6        | 128/781 [00:06<00:35, 18.52it/s]
 17%|#6        | 130/781 [00:07<00:35, 18.48it/s]
 17%|#6        | 132/781 [00:07<00:35, 18.38it/s]
 17%|#7        | 134/781 [00:07<00:35, 18.39it/s]
 17%|#7        | 136/781 [00:07<00:35, 18.34it/s]
 18%|#7        | 138/781 [00:07<00:34, 18.37it/s]
 18%|#7        | 140/781 [00:07<00:34, 18.38it/s]
 18%|#8        | 142/781 [00:07<00:34, 18.40it/s]
 18%|#8        | 144/781 [00:07<00:34, 18.41it/s]
 19%|#8        | 146/781 [00:07<00:34, 18.40it/s]
 19%|#8        | 148/781 [00:08<00:34, 18.40it/s]
 19%|#9        | 150/781 [00:08<00:34, 18.39it/s]
 19%|#9        | 152/781 [00:08<00:34, 18.42it/s]
 20%|#9        | 154/781 [00:08<00:34, 18.40it/s]
 20%|#9        | 156/781 [00:08<00:33, 18.39it/s]
 20%|##        | 158/781 [00:08<00:33, 18.38it/s]
 20%|##        | 160/781 [00:08<00:33, 18.37it/s]
 21%|##        | 162/781 [00:08<00:33, 18.38it/s]
 21%|##        | 164/781 [00:08<00:33, 18.41it/s]
 21%|##1       | 166/781 [00:09<00:33, 18.42it/s]
 22%|##1       | 168/781 [00:09<00:33, 18.45it/s]
 22%|##1       | 170/781 [00:09<00:33, 18.45it/s]
 22%|##2       | 172/781 [00:09<00:33, 18.45it/s]
 22%|##2       | 174/781 [00:09<00:32, 18.47it/s]
 23%|##2       | 176/781 [00:09<00:32, 18.45it/s]
 23%|##2       | 178/781 [00:09<00:32, 18.46it/s]
 23%|##3       | 180/781 [00:09<00:32, 18.55it/s]
 23%|##3       | 182/781 [00:09<00:32, 18.69it/s]
 24%|##3       | 184/781 [00:10<00:31, 18.75it/s]
 24%|##3       | 186/781 [00:10<00:31, 18.82it/s]
 24%|##4       | 188/781 [00:10<00:31, 18.86it/s]
 24%|##4       | 190/781 [00:10<00:31, 18.90it/s]
 25%|##4       | 192/781 [00:10<00:31, 18.90it/s]
 25%|##4       | 194/781 [00:10<00:31, 18.92it/s]
 25%|##5       | 196/781 [00:10<00:30, 18.92it/s]
 25%|##5       | 198/781 [00:10<00:30, 18.97it/s]
 26%|##5       | 200/781 [00:10<00:30, 18.98it/s]
 26%|##5       | 202/781 [00:10<00:30, 18.98it/s]
 26%|##6       | 204/781 [00:11<00:30, 18.95it/s]
 26%|##6       | 206/781 [00:11<00:30, 18.95it/s]
 27%|##6       | 208/781 [00:11<00:30, 18.92it/s]
 27%|##6       | 210/781 [00:11<00:30, 18.80it/s]
 27%|##7       | 212/781 [00:11<00:30, 18.70it/s]
 27%|##7       | 214/781 [00:11<00:30, 18.63it/s]
 28%|##7       | 216/781 [00:11<00:30, 18.59it/s]
 28%|##7       | 218/781 [00:11<00:30, 18.56it/s]
 28%|##8       | 220/781 [00:11<00:30, 18.52it/s]
 28%|##8       | 222/781 [00:12<00:30, 18.54it/s]
 29%|##8       | 224/781 [00:12<00:30, 18.52it/s]
 29%|##8       | 226/781 [00:12<00:29, 18.50it/s]
 29%|##9       | 228/781 [00:12<00:29, 18.47it/s]
 29%|##9       | 230/781 [00:12<00:29, 18.48it/s]
 30%|##9       | 232/781 [00:12<00:29, 18.47it/s]
 30%|##9       | 234/781 [00:12<00:29, 18.46it/s]
 30%|###       | 236/781 [00:12<00:29, 18.45it/s]
 30%|###       | 238/781 [00:12<00:29, 18.42it/s]
 31%|###       | 240/781 [00:13<00:29, 18.42it/s]
 31%|###       | 242/781 [00:13<00:29, 18.44it/s]
 31%|###1      | 244/781 [00:13<00:29, 18.48it/s]
 31%|###1      | 246/781 [00:13<00:29, 18.00it/s]
 32%|###1      | 248/781 [00:13<00:29, 18.11it/s]
 32%|###2      | 250/781 [00:13<00:29, 18.20it/s]
 32%|###2      | 252/781 [00:13<00:28, 18.28it/s]
 33%|###2      | 254/781 [00:13<00:28, 18.32it/s]
 33%|###2      | 256/781 [00:13<00:28, 18.32it/s]
 33%|###3      | 258/781 [00:14<00:28, 18.35it/s]
 33%|###3      | 260/781 [00:14<00:28, 18.41it/s]
 34%|###3      | 262/781 [00:14<00:28, 18.39it/s]
 34%|###3      | 264/781 [00:14<00:28, 18.41it/s]
 34%|###4      | 266/781 [00:14<00:27, 18.46it/s]
 34%|###4      | 268/781 [00:14<00:28, 18.18it/s]
 35%|###4      | 270/781 [00:14<00:27, 18.28it/s]
 35%|###4      | 272/781 [00:14<00:27, 18.34it/s]
 35%|###5      | 274/781 [00:14<00:27, 18.40it/s]
 35%|###5      | 276/781 [00:14<00:27, 18.53it/s]
 36%|###5      | 278/781 [00:15<00:27, 18.53it/s]
 36%|###5      | 280/781 [00:15<00:26, 18.60it/s]
 36%|###6      | 282/781 [00:15<00:26, 18.61it/s]
 36%|###6      | 284/781 [00:15<00:26, 18.62it/s]
 37%|###6      | 286/781 [00:15<00:26, 18.46it/s]
 37%|###6      | 288/781 [00:15<00:26, 18.57it/s]
 37%|###7      | 290/781 [00:15<00:26, 18.54it/s]
 37%|###7      | 292/781 [00:15<00:26, 18.27it/s]
 38%|###7      | 294/781 [00:15<00:26, 18.42it/s]
 38%|###7      | 296/781 [00:16<00:26, 18.38it/s]
 38%|###8      | 298/781 [00:16<00:26, 18.50it/s]
 38%|###8      | 300/781 [00:16<00:25, 18.54it/s]
 39%|###8      | 302/781 [00:16<00:25, 18.58it/s]
 39%|###8      | 304/781 [00:16<00:25, 18.59it/s]
 39%|###9      | 306/781 [00:16<00:25, 18.54it/s]
 39%|###9      | 308/781 [00:16<00:25, 18.55it/s]
 40%|###9      | 310/781 [00:16<00:25, 18.40it/s]
 40%|###9      | 312/781 [00:16<00:25, 18.74it/s]
 40%|####      | 314/781 [00:17<00:24, 18.86it/s]
 40%|####      | 316/781 [00:17<00:24, 18.94it/s]
 41%|####      | 318/781 [00:17<00:24, 18.94it/s]
 41%|####      | 320/781 [00:17<00:24, 18.99it/s]
 41%|####1     | 322/781 [00:17<00:24, 19.04it/s]
 41%|####1     | 324/781 [00:17<00:24, 19.02it/s]
 42%|####1     | 326/781 [00:17<00:23, 19.07it/s]
 42%|####1     | 328/781 [00:17<00:23, 19.01it/s]
 42%|####2     | 330/781 [00:17<00:23, 19.03it/s]
 43%|####2     | 332/781 [00:17<00:23, 19.04it/s]
 43%|####2     | 334/781 [00:18<00:23, 19.03it/s]
 43%|####3     | 336/781 [00:18<00:23, 19.07it/s]
 43%|####3     | 338/781 [00:18<00:23, 18.85it/s]
 44%|####3     | 340/781 [00:18<00:23, 18.91it/s]
 44%|####3     | 342/781 [00:18<00:23, 18.98it/s]
 44%|####4     | 344/781 [00:18<00:23, 18.97it/s]
 44%|####4     | 346/781 [00:18<00:22, 18.96it/s]
 45%|####4     | 348/781 [00:18<00:22, 18.98it/s]
 45%|####4     | 350/781 [00:18<00:22, 18.93it/s]
 45%|####5     | 352/781 [00:19<00:22, 18.83it/s]
 45%|####5     | 354/781 [00:19<00:22, 19.04it/s]
 46%|####5     | 356/781 [00:19<00:22, 19.09it/s]
 46%|####5     | 358/781 [00:19<00:22, 18.96it/s]
 46%|####6     | 360/781 [00:19<00:21, 19.18it/s]
 46%|####6     | 362/781 [00:19<00:21, 19.15it/s]
 47%|####6     | 364/781 [00:19<00:21, 19.16it/s]
 47%|####6     | 366/781 [00:19<00:21, 19.03it/s]
 47%|####7     | 368/781 [00:19<00:21, 19.24it/s]
 47%|####7     | 370/781 [00:19<00:21, 19.21it/s]
 48%|####7     | 372/781 [00:20<00:21, 19.23it/s]
 48%|####7     | 374/781 [00:20<00:21, 19.22it/s]
 48%|####8     | 376/781 [00:20<00:21, 19.22it/s]
 48%|####8     | 378/781 [00:20<00:20, 19.20it/s]
 49%|####8     | 380/781 [00:20<00:20, 19.18it/s]
 49%|####8     | 382/781 [00:20<00:20, 19.14it/s]
 49%|####9     | 384/781 [00:20<00:20, 19.16it/s]
 49%|####9     | 386/781 [00:20<00:20, 19.13it/s]
 50%|####9     | 388/781 [00:20<00:20, 18.96it/s]
 50%|####9     | 390/781 [00:21<00:20, 18.87it/s]
 50%|#####     | 392/781 [00:21<00:20, 18.78it/s]
 50%|#####     | 394/781 [00:21<00:20, 18.74it/s]
 51%|#####     | 396/781 [00:21<00:20, 18.56it/s]
 51%|#####     | 398/781 [00:21<00:20, 18.82it/s]
 51%|#####1    | 400/781 [00:21<00:20, 18.89it/s]
 51%|#####1    | 402/781 [00:21<00:19, 18.97it/s]
 52%|#####1    | 404/781 [00:21<00:19, 19.02it/s]
 52%|#####1    | 406/781 [00:21<00:19, 18.89it/s]
 52%|#####2    | 408/781 [00:21<00:19, 19.10it/s]
 52%|#####2    | 410/781 [00:22<00:19, 19.14it/s]
 53%|#####2    | 412/781 [00:22<00:19, 19.14it/s]
 53%|#####3    | 414/781 [00:22<00:19, 19.19it/s]
 53%|#####3    | 416/781 [00:22<00:19, 18.98it/s]
 54%|#####3    | 418/781 [00:22<00:18, 19.16it/s]
 54%|#####3    | 420/781 [00:22<00:18, 19.20it/s]
 54%|#####4    | 422/781 [00:22<00:18, 19.22it/s]
 54%|#####4    | 424/781 [00:22<00:18, 19.21it/s]
 55%|#####4    | 426/781 [00:22<00:18, 19.10it/s]
 55%|#####4    | 428/781 [00:23<00:18, 18.80it/s]
 55%|#####5    | 430/781 [00:23<00:18, 18.93it/s]
 55%|#####5    | 432/781 [00:23<00:18, 18.63it/s]
 56%|#####5    | 434/781 [00:23<00:18, 18.62it/s]
 56%|#####5    | 436/781 [00:23<00:18, 18.62it/s]
 56%|#####6    | 438/781 [00:23<00:18, 18.60it/s]
 56%|#####6    | 440/781 [00:23<00:18, 18.59it/s]
 57%|#####6    | 442/781 [00:23<00:18, 18.56it/s]
 57%|#####6    | 444/781 [00:23<00:18, 18.54it/s]
 57%|#####7    | 446/781 [00:23<00:18, 18.60it/s]
 57%|#####7    | 448/781 [00:24<00:17, 18.57it/s]
 58%|#####7    | 450/781 [00:24<00:17, 18.60it/s]
 58%|#####7    | 452/781 [00:24<00:17, 18.58it/s]
 58%|#####8    | 454/781 [00:24<00:17, 18.46it/s]
 58%|#####8    | 456/781 [00:24<00:17, 18.64it/s]
 59%|#####8    | 458/781 [00:24<00:17, 18.55it/s]
 59%|#####8    | 460/781 [00:24<00:17, 18.56it/s]
 59%|#####9    | 462/781 [00:24<00:17, 18.67it/s]
 59%|#####9    | 464/781 [00:24<00:16, 18.66it/s]
 60%|#####9    | 466/781 [00:25<00:16, 18.88it/s]
 60%|#####9    | 468/781 [00:25<00:16, 18.77it/s]
 60%|######    | 470/781 [00:25<00:16, 18.74it/s]
 60%|######    | 472/781 [00:25<00:16, 18.94it/s]
 61%|######    | 474/781 [00:25<00:16, 18.89it/s]
 61%|######    | 476/781 [00:25<00:16, 18.78it/s]
 61%|######1   | 478/781 [00:25<00:16, 18.84it/s]
 61%|######1   | 480/781 [00:25<00:16, 18.61it/s]
 62%|######1   | 482/781 [00:25<00:16, 18.68it/s]
 62%|######1   | 484/781 [00:26<00:15, 18.71it/s]
 62%|######2   | 486/781 [00:26<00:15, 18.65it/s]
 62%|######2   | 488/781 [00:26<00:15, 18.88it/s]
 63%|######2   | 490/781 [00:26<00:15, 18.81it/s]
 63%|######2   | 492/781 [00:26<00:15, 18.92it/s]
 63%|######3   | 494/781 [00:26<00:15, 18.93it/s]
 64%|######3   | 496/781 [00:26<00:15, 18.79it/s]
 64%|######3   | 498/781 [00:26<00:14, 18.88it/s]
 64%|######4   | 500/781 [00:26<00:14, 18.95it/s]
 64%|######4   | 502/781 [00:26<00:14, 18.98it/s]
 65%|######4   | 504/781 [00:27<00:14, 19.00it/s]
 65%|######4   | 506/781 [00:27<00:14, 19.02it/s]
 65%|######5   | 508/781 [00:27<00:14, 19.03it/s]
 65%|######5   | 510/781 [00:27<00:14, 19.01it/s]
 66%|######5   | 512/781 [00:27<00:14, 18.97it/s]
 66%|######5   | 514/781 [00:27<00:14, 19.02it/s]
 66%|######6   | 516/781 [00:27<00:13, 19.16it/s]
 66%|######6   | 518/781 [00:27<00:13, 19.08it/s]
 67%|######6   | 520/781 [00:27<00:13, 19.01it/s]
 67%|######6   | 522/781 [00:28<00:13, 18.88it/s]
 67%|######7   | 524/781 [00:28<00:13, 18.88it/s]
 67%|######7   | 526/781 [00:28<00:13, 18.92it/s]
 68%|######7   | 528/781 [00:28<00:13, 18.98it/s]
 68%|######7   | 530/781 [00:28<00:13, 18.96it/s]
 68%|######8   | 532/781 [00:28<00:13, 18.99it/s]
 68%|######8   | 534/781 [00:28<00:13, 18.98it/s]
 69%|######8   | 536/781 [00:28<00:12, 19.13it/s]
 69%|######8   | 538/781 [00:28<00:12, 19.00it/s]
 69%|######9   | 540/781 [00:28<00:12, 18.76it/s]
 69%|######9   | 542/781 [00:29<00:12, 18.69it/s]
 70%|######9   | 544/781 [00:29<00:12, 18.71it/s]
 70%|######9   | 546/781 [00:29<00:12, 18.58it/s]
 70%|#######   | 548/781 [00:29<00:12, 18.62it/s]
 70%|#######   | 550/781 [00:29<00:12, 18.52it/s]
 71%|#######   | 552/781 [00:29<00:12, 18.33it/s]
 71%|#######   | 554/781 [00:29<00:12, 18.43it/s]
 71%|#######1  | 556/781 [00:29<00:12, 18.47it/s]
 71%|#######1  | 558/781 [00:29<00:11, 18.66it/s]
 72%|#######1  | 560/781 [00:30<00:11, 18.77it/s]
 72%|#######1  | 562/781 [00:30<00:11, 18.86it/s]
 72%|#######2  | 564/781 [00:30<00:11, 18.92it/s]
 72%|#######2  | 566/781 [00:30<00:11, 18.96it/s]
 73%|#######2  | 568/781 [00:30<00:11, 19.12it/s]
 73%|#######2  | 570/781 [00:30<00:11, 19.06it/s]
 73%|#######3  | 572/781 [00:30<00:11, 18.89it/s]
 73%|#######3  | 574/781 [00:30<00:10, 18.91it/s]
 74%|#######3  | 576/781 [00:30<00:10, 19.05it/s]
 74%|#######4  | 578/781 [00:31<00:10, 18.99it/s]
 74%|#######4  | 580/781 [00:31<00:10, 18.82it/s]
 75%|#######4  | 582/781 [00:31<00:10, 18.96it/s]
 75%|#######4  | 584/781 [00:31<00:10, 18.84it/s]
 75%|#######5  | 586/781 [00:31<00:10, 18.87it/s]
 75%|#######5  | 588/781 [00:31<00:10, 18.93it/s]
 76%|#######5  | 590/781 [00:31<00:10, 18.97it/s]
 76%|#######5  | 592/781 [00:31<00:09, 19.15it/s]
 76%|#######6  | 594/781 [00:31<00:09, 19.05it/s]
 76%|#######6  | 596/781 [00:31<00:09, 18.89it/s]
 77%|#######6  | 598/781 [00:32<00:09, 19.00it/s]
 77%|#######6  | 600/781 [00:32<00:09, 18.86it/s]
 77%|#######7  | 602/781 [00:32<00:09, 18.92it/s]
 77%|#######7  | 604/781 [00:32<00:09, 18.89it/s]
 78%|#######7  | 606/781 [00:32<00:09, 19.05it/s]
 78%|#######7  | 608/781 [00:32<00:09, 18.69it/s]
 78%|#######8  | 610/781 [00:32<00:09, 18.83it/s]
 78%|#######8  | 612/781 [00:32<00:08, 18.92it/s]
 79%|#######8  | 614/781 [00:32<00:08, 18.96it/s]
 79%|#######8  | 616/781 [00:33<00:08, 18.99it/s]
 79%|#######9  | 618/781 [00:33<00:08, 19.06it/s]
 79%|#######9  | 620/781 [00:33<00:08, 19.08it/s]
 80%|#######9  | 622/781 [00:33<00:08, 19.10it/s]
 80%|#######9  | 624/781 [00:33<00:08, 18.98it/s]
 80%|########  | 626/781 [00:33<00:08, 18.92it/s]
 80%|########  | 628/781 [00:33<00:08, 18.89it/s]
 81%|########  | 630/781 [00:33<00:08, 18.86it/s]
 81%|########  | 632/781 [00:33<00:07, 18.89it/s]
 81%|########1 | 634/781 [00:33<00:07, 18.82it/s]
 81%|########1 | 636/781 [00:34<00:07, 18.80it/s]
 82%|########1 | 638/781 [00:34<00:07, 18.81it/s]
 82%|########1 | 640/781 [00:34<00:07, 18.85it/s]
 82%|########2 | 642/781 [00:34<00:07, 18.83it/s]
 82%|########2 | 644/781 [00:34<00:07, 18.94it/s]
 83%|########2 | 646/781 [00:34<00:07, 19.08it/s]
 83%|########2 | 648/781 [00:34<00:06, 19.13it/s]
 83%|########3 | 650/781 [00:34<00:06, 19.00it/s]
 83%|########3 | 652/781 [00:34<00:06, 18.89it/s]
 84%|########3 | 654/781 [00:35<00:06, 18.93it/s]
 84%|########3 | 656/781 [00:35<00:06, 18.99it/s]
 84%|########4 | 658/781 [00:35<00:06, 19.00it/s]
 85%|########4 | 660/781 [00:35<00:06, 19.04it/s]
 85%|########4 | 662/781 [00:35<00:06, 19.02it/s]
 85%|########5 | 664/781 [00:35<00:06, 19.01it/s]
 85%|########5 | 666/781 [00:35<00:06, 19.05it/s]
 86%|########5 | 668/781 [00:35<00:05, 18.99it/s]
 86%|########5 | 670/781 [00:35<00:05, 19.01it/s]
 86%|########6 | 672/781 [00:35<00:05, 18.94it/s]
 86%|########6 | 674/781 [00:36<00:05, 18.96it/s]
 87%|########6 | 676/781 [00:36<00:05, 18.98it/s]
 87%|########6 | 678/781 [00:36<00:05, 19.02it/s]
 87%|########7 | 680/781 [00:36<00:05, 18.89it/s]
 87%|########7 | 682/781 [00:36<00:05, 18.86it/s]
 88%|########7 | 684/781 [00:36<00:05, 18.89it/s]
 88%|########7 | 686/781 [00:36<00:04, 19.02it/s]
 88%|########8 | 688/781 [00:36<00:04, 19.01it/s]
 88%|########8 | 690/781 [00:36<00:04, 19.04it/s]
 89%|########8 | 692/781 [00:37<00:04, 19.08it/s]
 89%|########8 | 694/781 [00:37<00:04, 19.14it/s]
 89%|########9 | 696/781 [00:37<00:04, 19.18it/s]
 89%|########9 | 698/781 [00:37<00:04, 19.20it/s]
 90%|########9 | 700/781 [00:37<00:04, 19.23it/s]
 90%|########9 | 702/781 [00:37<00:04, 19.26it/s]
 90%|######### | 704/781 [00:37<00:04, 19.24it/s]
 90%|######### | 706/781 [00:37<00:03, 19.26it/s]
 91%|######### | 708/781 [00:37<00:03, 19.25it/s]
 91%|######### | 710/781 [00:37<00:03, 19.26it/s]
 91%|#########1| 712/781 [00:38<00:03, 19.22it/s]
 91%|#########1| 714/781 [00:38<00:03, 19.24it/s]
 92%|#########1| 716/781 [00:38<00:03, 19.22it/s]
 92%|#########1| 718/781 [00:38<00:03, 19.25it/s]
 92%|#########2| 720/781 [00:38<00:03, 19.21it/s]
 92%|#########2| 722/781 [00:38<00:03, 19.20it/s]
 93%|#########2| 724/781 [00:38<00:02, 19.21it/s]
 93%|#########2| 726/781 [00:38<00:02, 19.27it/s]
 93%|#########3| 728/781 [00:38<00:02, 19.21it/s]
 93%|#########3| 730/781 [00:38<00:02, 19.08it/s]
 94%|#########3| 732/781 [00:39<00:02, 18.99it/s]
 94%|#########3| 734/781 [00:39<00:02, 18.94it/s]
 94%|#########4| 736/781 [00:39<00:02, 18.95it/s]
 94%|#########4| 738/781 [00:39<00:02, 19.07it/s]
 95%|#########4| 740/781 [00:39<00:02, 19.16it/s]
 95%|#########5| 742/781 [00:39<00:02, 19.22it/s]
 95%|#########5| 744/781 [00:39<00:01, 19.27it/s]
 96%|#########5| 746/781 [00:39<00:01, 19.26it/s]
 96%|#########5| 748/781 [00:39<00:01, 19.30it/s]
 96%|#########6| 750/781 [00:40<00:01, 19.32it/s]
 96%|#########6| 752/781 [00:40<00:01, 19.32it/s]
 97%|#########6| 754/781 [00:40<00:01, 19.30it/s]
 97%|#########6| 756/781 [00:40<00:01, 19.29it/s]
 97%|#########7| 758/781 [00:40<00:01, 19.32it/s]
 97%|#########7| 760/781 [00:40<00:01, 19.32it/s]
 98%|#########7| 762/781 [00:40<00:00, 19.33it/s]
 98%|#########7| 764/781 [00:40<00:00, 19.32it/s]
 98%|#########8| 766/781 [00:40<00:00, 19.33it/s]
 98%|#########8| 768/781 [00:40<00:00, 19.31it/s]
 99%|#########8| 770/781 [00:41<00:00, 19.25it/s]
 99%|#########8| 772/781 [00:41<00:00, 19.24it/s]
 99%|#########9| 774/781 [00:41<00:00, 19.23it/s]
 99%|#########9| 776/781 [00:41<00:00, 19.24it/s]
100%|#########9| 778/781 [00:41<00:00, 19.24it/s]
100%|#########9| 780/781 [00:41<00:00, 19.26it/s]
100%|##########| 781/781 [00:41<00:00, 18.75it/s]






=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================







=======================================================   RESTART [11-27 02:00:37]   =======================================================
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/utils/misc.py:354: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  ckpt = torch.load(all_ckpt[0], map_location='cpu')
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2198: UserWarning: Overwriting vit_tiny_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2207: UserWarning: Overwriting vit_tiny_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_tiny_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_tiny_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2216: UserWarning: Overwriting vit_small_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2225: UserWarning: Overwriting vit_small_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2234: UserWarning: Overwriting vit_small_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2243: UserWarning: Overwriting vit_small_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2252: UserWarning: Overwriting vit_small_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2261: UserWarning: Overwriting vit_base_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2271: UserWarning: Overwriting vit_base_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2281: UserWarning: Overwriting vit_base_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2291: UserWarning: Overwriting vit_base_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2301: UserWarning: Overwriting vit_base_patch8_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch8_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch8_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2311: UserWarning: Overwriting vit_large_patch32_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2320: UserWarning: Overwriting vit_large_patch32_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch32_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch32_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2330: UserWarning: Overwriting vit_large_patch16_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2340: UserWarning: Overwriting vit_large_patch16_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2350: UserWarning: Overwriting vit_large_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2359: UserWarning: Overwriting vit_huge_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2368: UserWarning: Overwriting vit_giant_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2377: UserWarning: Overwriting vit_gigantic_patch14_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2387: UserWarning: Overwriting vit_base_patch16_224_miil in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_224_miil. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_224_miil(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2398: UserWarning: Overwriting vit_medium_patch16_gap_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2410: UserWarning: Overwriting vit_medium_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2422: UserWarning: Overwriting vit_medium_patch16_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2434: UserWarning: Overwriting vit_betwixt_patch16_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2446: UserWarning: Overwriting vit_base_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2457: UserWarning: Overwriting vit_huge_patch14_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2468: UserWarning: Overwriting vit_huge_patch16_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch16_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch16_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2479: UserWarning: Overwriting vit_giant_patch16_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch16_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch16_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2491: UserWarning: Overwriting vit_xsmall_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_xsmall_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_xsmall_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2500: UserWarning: Overwriting vit_medium_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2510: UserWarning: Overwriting vit_medium_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2519: UserWarning: Overwriting vit_betwixt_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2529: UserWarning: Overwriting vit_base_patch32_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2540: UserWarning: Overwriting vit_base_patch32_clip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2551: UserWarning: Overwriting vit_base_patch32_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2562: UserWarning: Overwriting vit_base_patch32_clip_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2573: UserWarning: Overwriting vit_base_patch16_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2583: UserWarning: Overwriting vit_base_patch16_clip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2593: UserWarning: Overwriting vit_large_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2603: UserWarning: Overwriting vit_large_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2613: UserWarning: Overwriting vit_huge_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2623: UserWarning: Overwriting vit_huge_patch14_clip_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2633: UserWarning: Overwriting vit_huge_patch14_clip_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2643: UserWarning: Overwriting vit_giant_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2656: UserWarning: Overwriting vit_gigantic_patch14_clip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_gigantic_patch14_clip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_gigantic_patch14_clip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2669: UserWarning: Overwriting vit_base_patch32_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2681: UserWarning: Overwriting vit_base_patch16_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2693: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2705: UserWarning: Overwriting vit_large_patch14_clip_quickgelu_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_clip_quickgelu_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_clip_quickgelu_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2717: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2729: UserWarning: Overwriting vit_huge_patch14_clip_quickgelu_378 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_clip_quickgelu_378. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_clip_quickgelu_378(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2743: UserWarning: Overwriting vit_base_patch32_plus_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch32_plus_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch32_plus_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2753: UserWarning: Overwriting vit_base_patch16_plus_240 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_plus_240. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_plus_240(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2763: UserWarning: Overwriting vit_base_patch16_rpn_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_rpn_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_rpn_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2775: UserWarning: Overwriting vit_small_patch16_36x1_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_36x1_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_36x1_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2787: UserWarning: Overwriting vit_small_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2800: UserWarning: Overwriting vit_base_patch16_18x2_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_18x2_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_18x2_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2812: UserWarning: Overwriting eva_large_patch14_196 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_196. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_196(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2821: UserWarning: Overwriting eva_large_patch14_336 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.eva_large_patch14_336. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def eva_large_patch14_336(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2829: UserWarning: Overwriting flexivit_small in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_small(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2838: UserWarning: Overwriting flexivit_base in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_base(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2847: UserWarning: Overwriting flexivit_large in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.flexivit_large. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def flexivit_large(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2856: UserWarning: Overwriting vit_base_patch16_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2869: UserWarning: Overwriting vit_large_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2882: UserWarning: Overwriting vit_huge_patch14_xp_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_huge_patch14_xp_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_huge_patch14_xp_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2895: UserWarning: Overwriting vit_small_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2905: UserWarning: Overwriting vit_base_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2915: UserWarning: Overwriting vit_large_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2925: UserWarning: Overwriting vit_giant_patch14_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2942: UserWarning: Overwriting vit_small_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_small_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_small_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2955: UserWarning: Overwriting vit_base_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2968: UserWarning: Overwriting vit_large_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2981: UserWarning: Overwriting vit_giant_patch14_reg4_dinov2 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_giant_patch14_reg4_dinov2. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_giant_patch14_reg4_dinov2(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:2998: UserWarning: Overwriting vit_base_patch16_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3008: UserWarning: Overwriting vit_base_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3018: UserWarning: Overwriting vit_base_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3028: UserWarning: Overwriting vit_base_patch16_siglip_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3038: UserWarning: Overwriting vit_large_patch16_siglip_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3048: UserWarning: Overwriting vit_large_patch16_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3058: UserWarning: Overwriting vit_so400m_patch14_siglip_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3068: UserWarning: Overwriting vit_so400m_patch14_siglip_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3078: UserWarning: Overwriting vit_base_patch16_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3089: UserWarning: Overwriting vit_base_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3100: UserWarning: Overwriting vit_base_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3111: UserWarning: Overwriting vit_base_patch16_siglip_gap_512 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_siglip_gap_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_siglip_gap_512(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3122: UserWarning: Overwriting vit_large_patch16_siglip_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3133: UserWarning: Overwriting vit_large_patch16_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_large_patch16_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_large_patch16_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3144: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_224 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_224(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3156: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_384 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_384(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3168: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_448 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_448. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_448(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3180: UserWarning: Overwriting vit_so400m_patch14_siglip_gap_896 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so400m_patch14_siglip_gap_896. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so400m_patch14_siglip_gap_896(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3192: UserWarning: Overwriting vit_wee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_wee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_wee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3203: UserWarning: Overwriting vit_pwee_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_pwee_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_pwee_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3214: UserWarning: Overwriting vit_little_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3225: UserWarning: Overwriting vit_little_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_little_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_little_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3236: UserWarning: Overwriting vit_medium_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3247: UserWarning: Overwriting vit_medium_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_medium_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_medium_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3258: UserWarning: Overwriting vit_mediumd_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_mediumd_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_mediumd_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3269: UserWarning: Overwriting vit_betwixt_patch16_reg1_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg1_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg1_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3280: UserWarning: Overwriting vit_betwixt_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_betwixt_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_betwixt_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3291: UserWarning: Overwriting vit_base_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_base_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_base_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3302: UserWarning: Overwriting vit_so150m_patch16_reg4_map_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_map_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_map_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/tokenizer/tokenizer_image/dino_enc/vision_transformer.py:3313: UserWarning: Overwriting vit_so150m_patch16_reg4_gap_256 in registry with tokenizer.tokenizer_image.dino_enc.vision_transformer.vit_so150m_patch16_reg4_gap_256. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  def vit_so150m_patch16_reg4_gap_256(pretrained: bool = False, **kwargs) -> VisionTransformer:
/home/user/VAR/train.py:99: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  checkpoint = torch.load(vae_ckpt, map_location="cpu")
/home/user/VAR/utils/amp_sc.py:27: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  self.scaler = torch.cuda.amp.GradScaler(init_scale=2. ** 11, growth_interval=1000) if self.using_fp16_rather_bf16 else None # only fp16 needs a scaler
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
wandb: Currently logged in as: kaiqiu. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.18.3
wandb: Run data is saved locally in /home/user/VAR/wandb/run-20241126_101103-l4l5mxys
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run true-dream-566
wandb:  View project at https://wandb.ai/kaiqiu/VAR
wandb:  View run at https://wandb.ai/kaiqiu/VAR/runs/l4l5mxys