{
  "_name_or_path": "MCG-NJU/videomae-base",
  "architectures": [
    "VideoMAEForVideoClassification"
  ],
  "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,
  "id2label": {
    "0": "Covering something with something",
    "1": "Moving part of something",
    "2": "Moving something down",
    "3": "Moving something up",
    "4": "Pushing something from left to right",
    "5": "Pushing something from right to left",
    "6": "Putting something into something",
    "7": "Putting something next to something",
    "8": "Putting something on a surface",
    "9": "Something falling like a rock",
    "10": "Squeezing something",
    "11": "Taking one of many similar things on the table",
    "12": "Tearing something into two pieces",
    "13": "Throwing something",
    "14": "Throwing something in the air and catching it",
    "15": "Turning something upside down",
    "16": "Uncovering something"
  },
  "image_size": 224,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "Covering something with something": 0,
    "Moving part of something": 1,
    "Moving something down": 2,
    "Moving something up": 3,
    "Pushing something from left to right": 4,
    "Pushing something from right to left": 5,
    "Putting something into something": 6,
    "Putting something next to something": 7,
    "Putting something on a surface": 8,
    "Something falling like a rock": 9,
    "Squeezing something": 10,
    "Taking one of many similar things on the table": 11,
    "Tearing something into two pieces": 12,
    "Throwing something": 13,
    "Throwing something in the air and catching it": 14,
    "Turning something upside down": 15,
    "Uncovering something": 16
  },
  "layer_norm_eps": 1e-12,
  "model_type": "videomae",
  "norm_pix_loss": true,
  "num_attention_heads": 12,
  "num_channels": 3,
  "num_frames": 16,
  "num_hidden_layers": 12,
  "patch_size": 16,
  "problem_type": "single_label_classification",
  "qkv_bias": true,
  "torch_dtype": "float32",
  "transformers_version": "4.26.1",
  "tubelet_size": 2,
  "use_mean_pooling": false
}