Training in progress, step 200

config.json

@@ -0,0 +1,208 @@
+{
+  "architectures": [
+    "ViltForQuestionAnswering"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "donut",
+    "1": "chopsticks",
+    "2": "air",
+    "3": "jeep",
+    "4": "sky",
+    "5": "wine tasting",
+    "6": "style",
+    "7": "forest",
+    "8": "tired",
+    "9": "shadows",
+    "10": "red and blue",
+    "11": "no",
+    "12": "blue and white",
+    "13": "yes",
+    "14": "sun",
+    "15": "woman",
+    "16": "wall",
+    "17": "yellow",
+    "18": "3",
+    "19": "king",
+    "20": "resting",
+    "21": "out",
+    "22": "little girl",
+    "23": "7",
+    "24": "suv",
+    "25": "hat",
+    "26": "brown",
+    "27": "train",
+    "28": "table",
+    "29": "white and blue",
+    "30": "cup",
+    "31": "shadow",
+    "32": "white",
+    "33": "crossing",
+    "34": "watching",
+    "35": "crown",
+    "36": "queen",
+    "37": "red",
+    "38": "doughnut",
+    "39": "woods",
+    "40": "birthday",
+    "41": "walking",
+    "42": "0",
+    "43": "gray",
+    "44": "shrimp",
+    "45": "pink",
+    "46": "girl",
+    "47": "blonde",
+    "48": "fashion",
+    "49": "2",
+    "50": "4",
+    "51": "car",
+    "52": "right",
+    "53": "many",
+    "54": "chair",
+    "55": "full",
+    "56": "green",
+    "57": "dog",
+    "58": "8",
+    "59": "bedroom",
+    "60": "black",
+    "61": "park",
+    "62": "wine",
+    "63": "wedding",
+    "64": "5",
+    "65": "picnic table",
+    "66": "6",
+    "67": "white and black",
+    "68": "bus",
+    "69": "hawaii",
+    "70": "skateboard",
+    "71": "backpack",
+    "72": "ice cream",
+    "73": "10",
+    "74": "double",
+    "75": "down",
+    "76": "plastic",
+    "77": "they aren't",
+    "78": "boy",
+    "79": "at table",
+    "80": "cat",
+    "81": "human",
+    "82": "1",
+    "83": "leather",
+    "84": "beige",
+    "85": "not sure",
+    "86": "red and yellow",
+    "87": "outside"
+  },
+  "image_size": 384,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "0": 42,
+    "1": 82,
+    "10": 73,
+    "2": 49,
+    "3": 18,
+    "4": 50,
+    "5": 64,
+    "6": 66,
+    "7": 23,
+    "8": 58,
+    "air": 2,
+    "at table": 79,
+    "backpack": 71,
+    "bedroom": 59,
+    "beige": 84,
+    "birthday": 40,
+    "black": 60,
+    "blonde": 47,
+    "blue and white": 12,
+    "boy": 78,
+    "brown": 26,
+    "bus": 68,
+    "car": 51,
+    "cat": 80,
+    "chair": 54,
+    "chopsticks": 1,
+    "crossing": 33,
+    "crown": 35,
+    "cup": 30,
+    "dog": 57,
+    "donut": 0,
+    "double": 74,
+    "doughnut": 38,
+    "down": 75,
+    "fashion": 48,
+    "forest": 7,
+    "full": 55,
+    "girl": 46,
+    "gray": 43,
+    "green": 56,
+    "hat": 25,
+    "hawaii": 69,
+    "human": 81,
+    "ice cream": 72,
+    "jeep": 3,
+    "king": 19,
+    "leather": 83,
+    "little girl": 22,
+    "many": 53,
+    "no": 11,
+    "not sure": 85,
+    "out": 21,
+    "outside": 87,
+    "park": 61,
+    "picnic table": 65,
+    "pink": 45,
+    "plastic": 76,
+    "queen": 36,
+    "red": 37,
+    "red and blue": 10,
+    "red and yellow": 86,
+    "resting": 20,
+    "right": 52,
+    "shadow": 31,
+    "shadows": 9,
+    "shrimp": 44,
+    "skateboard": 70,
+    "sky": 4,
+    "style": 6,
+    "sun": 14,
+    "suv": 24,
+    "table": 28,
+    "they aren't": 77,
+    "tired": 8,
+    "train": 27,
+    "walking": 41,
+    "wall": 16,
+    "watching": 34,
+    "wedding": 63,
+    "white": 32,
+    "white and black": 67,
+    "white and blue": 29,
+    "wine": 62,
+    "wine tasting": 5,
+    "woman": 15,
+    "woods": 39,
+    "yellow": 17,
+    "yes": 13
+  },
+  "layer_norm_eps": 1e-12,
+  "max_image_length": -1,
+  "max_position_embeddings": 40,
+  "modality_type_vocab_size": 2,
+  "model_type": "vilt",
+  "num_attention_heads": 12,
+  "num_channels": 3,
+  "num_hidden_layers": 12,
+  "num_images": -1,
+  "patch_size": 32,
+  "qkv_bias": true,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.51.3",
+  "type_vocab_size": 2,
+  "vocab_size": 30522
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:384f8e579ec707f4a2bfb88a4351a695e7c61115f8272c3d9a3fde4a6a027abf
+size 451682896

preprocessor_config.json

@@ -0,0 +1,24 @@
+{
+  "do_normalize": true,
+  "do_pad": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "ViltImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "processor_class": "ViltProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "shortest_edge": 384
+  },
+  "size_divisor": 32
+}

runs/May07_18-45-36_af9f7ad9db60/events.out.tfevents.1746643608.af9f7ad9db60.6046.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:894484411cbf6b3df7a3c5b5d407d8db9e9d0e9dc952358b9491f3fca4021875
+size 9164

special_tokens_map.json

@@ -0,0 +1,37 @@
+{
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,57 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 40,
+  "pad_token": "[PAD]",
+  "processor_class": "ViltProcessor",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9686829b6003b796418e4db8a3cf4216d650da6ca719cc9c626d1d2da598012
+size 5304