{
  "_name_or_path": "models/movienet_mtask/MCG-NJU/videomae-base-finetuned-kinetics",
  "architectures": [
    "CustomModel"
  ],
  "attention_probs_dropout_prob": 0.0,
  "decoder_hidden_size": 384,
  "decoder_intermediate_size": 1536,
  "decoder_num_attention_heads": 6,
  "decoder_num_hidden_layers": 4,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "image_size": 224,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "layer_norm_eps": 1e-12,
  "model_type": "videomae",
  "movement_id2label": {
    "0": "Static",
    "1": "Motion",
    "2": "Pull",
    "3": "Push"
  },
  "movement_label2id": {
    "Motion": 1,
    "Pull": 2,
    "Push": 3,
    "Static": 0
  },
  "norm_pix_loss": true,
  "num_attention_heads": 12,
  "num_channels": 3,
  "num_frames": 16,
  "num_hidden_layers": 12,
  "patch_size": 16,
  "qkv_bias": true,
  "scale_id2label": {
    "0": "ECS",
    "1": "CS",
    "2": "MS",
    "3": "FS",
    "4": "LS"
  },
  "scale_label2id": {
    "CS": 1,
    "ECS": 0,
    "FS": 3,
    "LS": 4,
    "MS": 2
  },
  "torch_dtype": "float32",
  "transformers_version": "4.38.2",
  "tubelet_size": 2,
  "use_mean_pooling": true
}