| { | |
| "_attn_implementation_autoset": true, | |
| "answer_space": [ | |
| "circle", | |
| "green", | |
| "red", | |
| "gray", | |
| "yes", | |
| "teal", | |
| "black", | |
| "rectangle", | |
| "yellow", | |
| "triangle", | |
| "brown", | |
| "blue", | |
| "no" | |
| ], | |
| "base_model_name": "blip2", | |
| "classification_input_dim": 768, | |
| "dataset_name": "easy-vqa", | |
| "image_text_hidden_size": 256, | |
| "image_token_index": null, | |
| "initializer_factor": 1.0, | |
| "initializer_range": 0.02, | |
| "interm_dim": 1024, | |
| "model_type": "blip-2", | |
| "multi_class_classifier": true, | |
| "num_query_tokens": 32, | |
| "qformer_config": { | |
| "model_type": "blip_2_qformer" | |
| }, | |
| "text_config": { | |
| "model_type": "opt" | |
| }, | |
| "transformers_version": "4.46.1", | |
| "use_decoder_only_language_model": true, | |
| "vision_config": { | |
| "model_type": "blip_2_vision_model" | |
| } | |
| } | |