Visual Question Answering
Transformers
Safetensors
English
llava_llama
text-generation
mol
multimodal
custom-code
Instructions to use wjdghks950/vicuna_clip_patch with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use wjdghks950/vicuna_clip_patch with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "visual-question-answering" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # pip install "transformers<5.0.0" from transformers import pipeline pipe = pipeline("visual-question-answering", model="wjdghks950/vicuna_clip_patch")# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("wjdghks950/vicuna_clip_patch", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download source_config.json from wjdghks950/vicuna_clip_patch: direct link, hf CLI and curl.
- Browser
- Download file 1.94 kB
-
https://huggingface.co/wjdghks950/vicuna_clip_patch/resolve/main/source_config.json
- Command line
-
hf download hf://wjdghks950/vicuna_clip_patch/source_config.json
-
curl -L -o source_config.json https://huggingface.co/wjdghks950/vicuna_clip_patch/resolve/main/source_config.json
1.94 kB
| { | |
| "_name_or_path": "lmsys/vicuna-13b-v1.5", | |
| "architectures": [ | |
| "RouterLlavaLlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "attn_level": "patch_layer", | |
| "bos_token_id": 1, | |
| "cap_softmax": false, | |
| "cap_softmax_value": 0.06, | |
| "confidence_gate": false, | |
| "confidence_gate_slope": 10.0, | |
| "confidence_gate_threshold": 0.6, | |
| "eos_token_id": 2, | |
| "force_reserve_layer": true, | |
| "freeze_mm_mlp_adapter": false, | |
| "freeze_residual_scaling_coef": true, | |
| "hidden_act": "silu", | |
| "hidden_size": 5120, | |
| "image_aspect_ratio": "pad", | |
| "initializer_range": 0.02, | |
| "intermediate_size": 13824, | |
| "learnable_gate": true, | |
| "max_length": 4096, | |
| "max_position_embeddings": 4096, | |
| "mm_hidden_size": 1024, | |
| "mm_patch_merge_type": "flat", | |
| "mm_projector_lr": 2e-05, | |
| "mm_projector_type": "mlp2x_gelu", | |
| "mm_text_hidden_size": 768, | |
| "mm_use_im_patch_token": false, | |
| "mm_use_im_start_end": false, | |
| "mm_use_text_enc": true, | |
| "mm_vision_select_feature": "patch", | |
| "mm_vision_select_layer": -2, | |
| "mm_vision_tower": "openai/clip-vit-large-patch14-336", | |
| "model_type": "llava_llama", | |
| "num_attention_heads": 40, | |
| "num_hidden_layers": 40, | |
| "num_key_value_heads": 40, | |
| "pad_token_id": 0, | |
| "pretraining_tp": 1, | |
| "reserve_layer": -2, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 10000.0, | |
| "router_lr": 2e-05, | |
| "save_router_stats": true, | |
| "save_steps_router_stats": 5, | |
| "text_attention_dim": 1024, | |
| "text_attention_type": "qk", | |
| "text_conditioning": "attention", | |
| "tie_word_embeddings": false, | |
| "tokenizer_model_max_length": 2048, | |
| "tokenizer_padding_side": "right", | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.37.2", | |
| "tune_mm_mlp_adapter": false, | |
| "tune_residual_scaling_coef": false, | |
| "use_cache": true, | |
| "use_mm_proj": true, | |
| "use_text_conditioned_router": true, | |
| "use_text_only_gating": false, | |
| "vision_tower_delta": true, | |
| "vocab_size": 32000 | |
| } | |