{ "architectures": [ "Blip2VisionModelWithProjection" ], "dtype": "float32", "image_text_hidden_size": 256, "image_token_index": null, "initializer_factor": 1.0, "initializer_range": 0.02, "model_type": "blip-2", "num_query_tokens": 32, "qformer_config": { "attention_dropout": 0.1, "attention_probs_dropout_prob": 0.1, "bos_token_id": 0, "cross_attention_frequency": 2, "dropout": 0.1, "encoder_hidden_size": 32, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 32, "initializer_range": 0.02, "intermediate_size": 37, "layer_norm_eps": 1e-12, "max_position_embeddings": 512, "model_type": "blip_2_qformer", "num_attention_heads": 4, "num_hidden_layers": 2, "pad_token_id": 0, "projection_dim": 32, "use_qformer_text_input": true, "vocab_size": 1024 }, "text_config": { "_remove_final_layer_norm": false, "activation_function": "relu", "attention_dropout": 0.0, "bos_token_id": 2, "do_layer_norm_before": true, "dropout": 0.1, "enable_bias": true, "eos_token_id": 2, "ffn_dim": 3072, "hidden_size": 768, "init_std": 0.02, "layer_norm_elementwise_affine": true, "layerdrop": 0.0, "max_position_embeddings": 2048, "model_type": "opt", "num_attention_heads": 12, "num_hidden_layers": 12, "pad_token_id": 1, "tie_word_embeddings": true, "use_cache": true, "vocab_size": 1024, "word_embed_proj_dim": 768 }, "transformers_version": "5.15.0.dev0", "use_decoder_only_language_model": true, "vision_config": { "attention_dropout": 0.1, "dropout": 0.1, "hidden_act": "gelu", "hidden_size": 32, "image_size": 30, "initializer_range": 1e-10, "intermediate_size": 37, "layer_norm_eps": 1e-06, "model_type": "blip_2_vision_model", "num_attention_heads": 4, "num_channels": 3, "num_hidden_layers": 2, "patch_size": 2, "projection_dim": 32, "qkv_bias": true } }