| { | |
| "_name_or_path": "/data02/models/pix2struct-large", | |
| "architectures": [ | |
| "Pix2StructForConditionalGeneration" | |
| ], | |
| "decoder_start_token_id": 0, | |
| "eos_token_id": 1, | |
| "initializer_factor": 1.0, | |
| "initializer_range": 0.02, | |
| "is_encoder_decoder": true, | |
| "is_vqa": false, | |
| "model_type": "pix2struct", | |
| "pad_token_id": 0, | |
| "text_config": { | |
| "d_ff": 3968, | |
| "encoder_hidden_size": 1536, | |
| "hidden_size": 1536, | |
| "initializer_range": 0.02, | |
| "model_type": "pix2struct_text_model", | |
| "num_heads": 24, | |
| "num_layers": 18, | |
| "vocab_size": 50344 | |
| }, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.44.0", | |
| "vision_config": { | |
| "d_ff": 3968, | |
| "hidden_size": 1536, | |
| "initializer_range": 0.02, | |
| "layer_norm_bias": false, | |
| "model_type": "pix2struct_vision_model", | |
| "num_attention_heads": 24, | |
| "num_channels": 3, | |
| "num_hidden_layers": 18, | |
| "patch_size": 16, | |
| "projection_dim": 768 | |
| } | |
| } | |