{ "_name_or_path": ".cache/qwen2_5/3b_instruction", "alpha": 0.0, "architectures": [ "Qwen2ForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 151643, "cap_hidden_size": 896, "cap_vocab_size": 151936, "eos_token_id": 151645, "freeze_cap_mm_mlp_adapter": true, "freeze_mm_mlp_adapter": false, "hidden_act": "silu", "hidden_size": 2048, "hidden_state_layer": -9, "image_aspect_ratio": "pad", "image_grid_pinpoints": null, "image_mm_hidden_size": 1024, "image_mm_projector_type": "mlp2x_gelu", "image_mm_vision_tower": ".cache/clip-vit-large-patch14-336", "initializer_range": 0.02, "intermediate_size": 11008, "is_cap_branch": true, "is_pool_branch": true, "llm_projector_type": "mlp2x_gelu", "max_position_embeddings": 32768, "max_window_layers": 70, "mm_hidden_size": 1408, "mm_projector_lr": 2e-05, "mm_projector_type": "mlp2x_gelu", "mm_use_box_start_end": false, "mm_use_im_patch_token": false, "mm_use_im_start_end": false, "mm_vision_select_feature": "patch", "mm_vision_select_layer": -2, "mm_vision_tower": ".cache/InternVideo2-Stage2_1B-224p-f4/InternVideo2-stage2_1b-224p-f4.pt", "model_type": "VideoQwenVQA", "num_attention_heads": 16, "num_hidden_layers": 36, "num_key_value_heads": 2, "rms_norm_eps": 1e-06, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": true, "torch_dtype": "bfloat16", "transformers_version": "4.41.0", "use_cache": true, "use_mm_proj": true, "use_sliding_window": false, "vocab_size": 151936 }