{
  "architectures": [
    "VitPoseForPoseEstimation"
  ],
  "backbone_config": {
    "attention_probs_dropout_prob": 0.0,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.0,
    "hidden_size": 32,
    "image_size": [
      128,
      96
    ],
    "initializer_range": 0.02,
    "intermediate_size": 37,
    "layer_norm_eps": 1e-12,
    "mlp_ratio": 4,
    "model_type": "vitpose_backbone",
    "num_attention_heads": 4,
    "num_channels": 3,
    "num_experts": 1,
    "num_hidden_layers": 2,
    "out_features": [
      "stage2"
    ],
    "out_indices": [
      -1
    ],
    "part_features": 256,
    "patch_size": [
      8,
      8
    ],
    "qkv_bias": true,
    "stage_names": [
      "stem",
      "stage1",
      "stage2"
    ]
  },
  "dtype": "float32",
  "initializer_range": 0.02,
  "model_type": "vitpose",
  "scale_factor": 4,
  "transformers_version": "5.16.0.dev0",
  "use_simple_decoder": true
}