{ "architectures": [ "VideoMAEForVideoClassification" ], "attention_probs_dropout_prob": 0.1, "decoder_hidden_size": 384, "decoder_intermediate_size": 1536, "decoder_num_attention_heads": 6, "decoder_num_hidden_layers": 4, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 32, "image_size": 10, "initializer_range": 0.02, "intermediate_size": 37, "layer_norm_eps": 1e-12, "model_type": "videomae", "norm_pix_loss": true, "num_attention_heads": 4, "num_channels": 3, "num_frames": 2, "num_hidden_layers": 5, "patch_size": 2, "qkv_bias": true, "torch_dtype": "float32", "transformers_version": "4.28.0.dev0", "tubelet_size": 2, "use_mean_pooling": true }