| { |
| "architectures": [ |
| "blip3oQwenForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 151643, |
| "diffusion_name_or_path": "Efficient-Large-Model/SANA1.5_1.6B_1024px_diffusers", |
| "eos_token_id": 151645, |
| "head_dim": 128, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "image_aspect_ratio": "square", |
| "image_end_tag_id": 151670, |
| "image_end_token_id": 217209, |
| "image_start_tag_id": 151669, |
| "image_start_token_id": 151674, |
| "initializer_range": 0.02, |
| "intermediate_size": 6144, |
| "max_position_embeddings": 40960, |
| "max_window_layers": 28, |
| "mm_hidden_size": 1152, |
| "mm_patch_merge_type": "flat", |
| "mm_tunable_parts": "mm_language_model", |
| "mm_use_im_start_end": true, |
| "mm_vision_select_feature": "patch", |
| "mm_vision_select_layer": -2, |
| "mm_vision_tower": "/fsx/sfr/data/jiuhai/hub/models--csuhan--TA-Tok/snapshots/ac4dbb76a42d8b8bc92532b6fdb90c42e2d0a447/ta_tok.pth", |
| "mm_vision_tower_lr": null, |
| "model_type": "qwen3", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 28, |
| "num_image_tokens": 65536, |
| "num_key_value_heads": 8, |
| "num_new_tokens": 65536, |
| "num_scale_tokens": 3, |
| "pos_skipping_range": 4096, |
| "rms_norm_eps": 1e-06, |
| "rope_scaling": null, |
| "rope_theta": 1000000, |
| "scale_end_token_id": 151673, |
| "scale_start_token_id": 151671, |
| "sliding_window": null, |
| "tie_word_embeddings": true, |
| "tokenizer_model_max_length": 2048, |
| "tokenizer_padding_side": "right", |
| "torch_dtype": "bfloat16", |
| "transformers_version": "4.51.3", |
| "use_cache": true, |
| "use_mm_proj": true, |
| "use_pos_skipping": false, |
| "use_sliding_window": false, |
| "vision_tower_pretrained": null, |
| "vocab_size": 217210 |
| } |
|
|