{ "_name_or_path": "/9950backfile/yptang/projects/temp_method2_2v_temp_final/outputs/video_language/qwen2-vl-2B_Vit-G/stage1/2025-05-11-15-31-10_stage1_224_pool8w4_4w8_data-558k_webvid_second_type1_method_fusion_add_w-stage1", "architectures": [ "Qwen2VLForConditionalGeneration" ], "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151645, "hidden_act": "silu", "hidden_size": 1536, "image_also_pool": false, "image_aspect_ratio": "square", "image_grid_pinpoints": [ [ 224, 448 ], [ 448, 224 ], [ 448, 448 ], [ 672, 224 ], [ 224, 672 ] ], "image_token_id": 151655, "initializer_range": 0.02, "intermediate_size": 8960, "max_position_embeddings": 32768, "max_window_layers": 28, "mm_audio_select_layer": -1, "mm_audio_tower": "/public/zhaozijia/ckpt/whisper-large-v2", "mm_hidden_size": 1408, "mm_hidden_size_audio": 1280, "mm_patch_merge_type": "flat", "mm_projector_lr": null, "mm_projector_type": "mlp2x_gelu", "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": "/public/chensihan/projects/LLaMA-VID/model_zoo/LAVIS/eva_vit_g.pth", "model_type": "qwen2", "multagent_resolution": "128w4_32w8", "num_attention_heads": 12, "num_hidden_layers": 28, "num_key_value_heads": 2, "rms_norm_eps": 1e-06, "rope_scaling": null, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": true, "tokenizer_model_max_length": 4096, "tokenizer_padding_side": "right", "torch_dtype": "bfloat16", "transformers_version": "4.37.2", "tune_mm_mlp_adapter": false, "unfreeze_mm_audio_tower": false, "unfreeze_mm_vision_tower": false, "use_cache": true, "use_mm_proj": true, "use_multagent": "v0", "use_sliding_window": false, "use_temp_prompt": "second_type1", "video_pooling_size": 8, "video_token_id": 151656, "vision_config": { "depth": 32, "embed_dim": 1280, "hidden_size": 1536, "in_chans": 3, "mlp_ratio": 4, "num_heads": 16, "patch_size": 14, "spatial_merge_size": 2, "spatial_patch_size": 14, "temporal_patch_size": 2 }, "vision_end_token_id": 151653, "vision_start_token_id": 151652, "vision_token_id": 151654, "vision_tower_lr": null, "vocab_size": 151936 }