| { | |
| "architectures": [ | |
| "UniLIP_InternVLForCausalLM" | |
| ], | |
| "mllm_path":"OpenGVLab/InternVL3-1B", | |
| "mllm_hf_path":"OpenGVLab/InternVL3-1B-hf", | |
| "dit_path":"Efficient-Large-Model/Sana_600M_512px_diffusers", | |
| "vae_path":"mit-han-lab/dc-ae-f32c32-sana-1.1-diffusers", | |
| "unilip_factor":10.6, | |
| "connect_layer": 6, | |
| "downsample_ratio": 0.5, | |
| "freeze_mm_mlp_adapter": false, | |
| "gen_pooling": "early_pool2d_4", | |
| "gen_vision_tower": "eva-clip-E-14-plus", | |
| "image_aspect_ratio": "square", | |
| "image_seq_length": 256, | |
| "image_token_id": 151667, | |
| "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, | |
| "model_type": "unilip_internvl", | |
| "n_query": 256, | |
| "pad_token_id": 151643, | |
| "projector_hidden_act": "gelu", | |
| "text_config": { | |
| "architectures": [ | |
| "Qwen2ForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 151643, | |
| "eos_token_id": 151645, | |
| "hidden_act": "silu", | |
| "hidden_size": 896, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4864, | |
| "max_position_embeddings": 32768, | |
| "max_window_layers": 70, | |
| "model_type": "qwen2", | |
| "num_attention_heads": 14, | |
| "num_hidden_layers": 24, | |
| "num_key_value_heads": 2, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": { | |
| "factor": 2.0, | |
| "rope_type": "dynamic", | |
| "type": "dynamic" | |
| }, | |
| "rope_theta": 1000000.0, | |
| "sliding_window": null, | |
| "torch_dtype": "bfloat16", | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "vocab_size": 151678 | |
| }, | |
| "tokenizer_model_max_length": 1024, | |
| "tokenizer_padding_side": "right", | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.52.1", | |
| "tune_mm_mlp_adapter": false, | |
| "use_cache": false, | |
| "use_mm_proj": true, | |
| "vision_config": { | |
| "architectures": [ | |
| "InternVisionModel" | |
| ], | |
| "attention_bias": true, | |
| "attention_dropout": 0.0, | |
| "dropout": 0.0, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.0, | |
| "hidden_size": 1024, | |
| "image_size": [ | |
| 448, | |
| 448 | |
| ], | |
| "initializer_factor": 0.1, | |
| "initializer_range": 1e-10, | |
| "intermediate_size": 4096, | |
| "layer_norm_eps": 1e-06, | |
| "layer_scale_init_value": 0.1, | |
| "model_type": "internvl_vision", | |
| "norm_type": "layer_norm", | |
| "num_attention_heads": 16, | |
| "num_channels": 3, | |
| "num_hidden_layers": 24, | |
| "patch_size": [ | |
| 14, | |
| 14 | |
| ], | |
| "projection_dropout": 0.0, | |
| "torch_dtype": "bfloat16", | |
| "use_absolute_position_embeddings": true, | |
| "use_mask_token": false, | |
| "use_mean_pooling": true, | |
| "use_qk_norm": false | |
| }, | |
| "vision_feature_layer": -1, | |
| "vision_feature_select_strategy": "default", | |
| "vision_tower_pretrained": null | |
| } | |