Instructions to use shauray/ViTPose with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use shauray/ViTPose with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoImageProcessor, ViTPoseForPoseEstimation processor = AutoImageProcessor.from_pretrained("shauray/ViTPose") model = ViTPoseForPoseEstimation.from_pretrained("shauray/ViTPose", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download config.json from shauray/ViTPose: direct link, hf CLI and curl.
- Browser
- Download file 760 Bytes
-
https://huggingface.co/shauray/ViTPose/resolve/548d46c7aa824581b62ccb906be9e92d484cdaff/config.json
- Command line
-
hf download hf://shauray/ViTPose@548d46c7aa824581b62ccb906be9e92d484cdaff/config.json
-
curl -L -o config.json https://huggingface.co/shauray/ViTPose/resolve/548d46c7aa824581b62ccb906be9e92d484cdaff/config.json
760 Bytes
| { | |
| "architectures": [ | |
| "ViTPoseForPoseEstimation" | |
| ], | |
| "depth": 12, | |
| "drop_path_rate": 0.1, | |
| "dropout_p": 0.0, | |
| "embed_dim": 384, | |
| "flip_test": false, | |
| "img_size": [ | |
| 256, | |
| 192 | |
| ], | |
| "initializer_range": 1, | |
| "kernel": 3, | |
| "keypoint_in_channels": 382, | |
| "keypoint_num_deconv_filters": [ | |
| 256, | |
| 256 | |
| ], | |
| "keypoint_num_deconv_kernels": [ | |
| 4, | |
| 4 | |
| ], | |
| "keypoint_num_deconv_layer": 2, | |
| "mlp_ratio": 4, | |
| "model_type": "vitpose", | |
| "num_attention_heads": 12, | |
| "num_channels": 3, | |
| "num_joints": 17, | |
| "num_output_channels": 17, | |
| "patch_size": 16, | |
| "qkv_bias": true, | |
| "ratio": 1, | |
| "target_type": "GaussianHeatMap", | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.32.0.dev0", | |
| "udp": true, | |
| "use_checkpoint": false | |
| } |