model_final/config.json

{
  "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,
  "dtype": "float32",
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "id2label": {
    "0": "Clapping",
    "1": "Meet and Split",
    "2": "Sitting",
    "3": "Standing Still",
    "4": "Walking",
    "5": "Walking While Reading Book",
    "6": "Walking While Using Phone"
  },
  "image_size": 224,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "Clapping": 0,
    "Meet and Split": 1,
    "Sitting": 2,
    "Standing Still": 3,
    "Walking": 4,
    "Walking While Reading Book": 5,
    "Walking While Using Phone": 6
  },
  "layer_norm_eps": 1e-12,
  "model_type": "videomae",
  "norm_pix_loss": false,
  "num_attention_heads": 12,
  "num_channels": 3,
  "num_frames": 16,
  "num_hidden_layers": 12,
  "patch_size": 16,
  "qkv_bias": true,
  "transformers_version": "5.2.0",
  "tubelet_size": 2,
  "use_mean_pooling": true
}