{ "_name_or_path": "work_dirs/Steve_uitars_7B_128k_sft0218_fullcorrect_with_single_frame_from_grounding_hf", "architectures": [ "Qwen2VLForConditionalGeneration" ], "asr_align": false, "attention_dropout": 0.0, "bos_token_id": 151643, "check_data_modality": false, "compress": false, "compress_gap": 33, "down_factor_1": 1, "down_factor_2": 1, "eos_token_id": 151645, "freeze_mm_mlp_adapter": true, "hidden_act": "silu", "hidden_size": 3584, "image_aspect_ratio": "qwen2vl", "image_global": false, "image_grid": 1, "image_grid_pinpoints": null, "image_token_id": 151655, "initializer_range": 0.02, "intermediate_size": 18944, "keep_global": false, "keep_rate": 1.0, "max_position_embeddings": 128000, "max_window_layers": 28, "mm_hidden_size": 3584, "mm_projector_lr": null, "mm_projector_type": "identity", "mm_use_im_patch_token": false, "mm_use_im_start_end": false, "mm_use_mrope": false, "mm_vision_select_feature": "patch", "mm_vision_select_layer": -1, "mm_vision_tower": "model_zoo/vision/UI-TARS-7B-DPO-ViT-Qwen2VL", "model_type": "qwen2_vl", "num_attention_heads": 28, "num_hidden_layers": 28, "num_key_value_heads": 4, "overlap_ratio": 0.0, "rms_norm_eps": 1e-06, "rope_scaling": { "mrope_section": [ 16, 24, 24 ], "rope_type": "default", "type": "default" }, "rope_theta": 1000000.0, "sliding_window": 32768, "speech_overlap": false, "speech_overlap_time": 30, "tie_word_embeddings": false, "tokenizer_model_max_length": 128000, "tokenizer_padding_side": "right", "torch_dtype": "bfloat16", "train_modality": "text_image", "transformers_version": "4.46.0", "tune_mm_mlp_adapter": false, "use_cache": true, "use_mm_proj": true, "use_sliding_window": false, "video_token_id": 151656, "vision_config": { "in_chans": 3, "model_type": "qwen2_vl", "spatial_patch_size": 14 }, "vision_end_token_id": 151653, "vision_start_token_id": 151652, "vision_token_id": 151654, "vocab_size": 152064 }