| { | |
| "architectures": [ | |
| "VisualBertForPreTraining" | |
| ], | |
| "attention_probs_dropout_prob": 0.1, | |
| "bos_token_id": 0, | |
| "bypass_transformer": false, | |
| "eos_token_id": 2, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 3072, | |
| "layer_norm_eps": 1e-12, | |
| "max_position_embeddings": 512, | |
| "model_type": "visual_bert", | |
| "num_attention_heads": 12, | |
| "num_hidden_layers": 12, | |
| "pad_token_id": 1, | |
| "special_visual_initialize": true, | |
| "transformers_version": "4.7.0.dev0", | |
| "type_vocab_size": 2, | |
| "visual_embedding_dim": 2048, | |
| "vocab_size": 30522 | |
| } | |