{ "_name_or_path": "/mnt/data/user/tc_agi/zyp/llavafeatup/model_checkpoints/vicuna-13b-v1.5", "add_faster_video": false, "add_time_instruction": false, "architectures": [ "LlavaLlamaForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 1, "eos_token_id": 2, "faster_token_stride": 10, "feature_scale_mask": 5, "force_sample": false, "hidden_act": "silu", "hidden_size": 5120, "image_aspect_ratio": "square", "image_crop_resolution": null, "image_grid_pinpoints": null, "image_split_resolution": null, "initializer_range": 0.02, "intermediate_size": 13824, "jbu_ckpt": "/mnt/data/user/tc_agi/zyp/models/clip-large_jbu_4x_stack_cocostuff_attention_crf_0.001_tv_0.0_ent_0.0-0.001-False-30-2-1_2000.ckpt", "max_length": 4096, "max_position_embeddings": 4096, "mm_hidden_size": 1024, "mm_newline_position": "grid", "mm_patch_merge_type": "flat", "mm_projector_lr": null, "mm_projector_type": "adapt_spatial_resampler_v2", "mm_resampler_type": null, "mm_spatial_pool_mode": "bilinear", "mm_spatial_pool_stride": null, "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model", "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": "/mnt/data/user/tc_agi/zyp/featup/models/huggingface/hub/models--openai--clip-vit-large-patch14-336", "mm_vision_tower_lr": null, "model_mode": "uhd_v2", "model_type": "llama", "num_attention_heads": 40, "num_hidden_layers": 40, "num_key_value_heads": 40, "pad_token_id": 0, "pos_skipping_range": 4096, "pretraining_tp": 1, "rms_norm_eps": 1e-05, "rope_scaling": null, "rope_theta": 10000.0, "tie_word_embeddings": false, "tokenizer_model_max_length": 4096, "tokenizer_padding_side": "right", "torch_dtype": "bfloat16", "transformers_version": "4.40.1", "use_cache": true, "use_mm_proj": true, "use_pos_skipping": false, "vision_tower_pretrained": null, "vocab_size": 32000 }