Instructions to use wjdghks950/phi1_5_siglip_patch with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use wjdghks950/phi1_5_siglip_patch with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("microsoft/phi-1_5") model = PeftModel.from_pretrained(base_model, "wjdghks950/phi1_5_siglip_patch") - Notebooks
- Google Colab
- Kaggle
Download inference_config.json from wjdghks950/phi1_5_siglip_patch: direct link, hf CLI and curl.
- Browser
- Download file 549 Bytes
-
https://huggingface.co/wjdghks950/phi1_5_siglip_patch/resolve/main/inference_config.json
- Command line
-
hf download hf://wjdghks950/phi1_5_siglip_patch/inference_config.json
-
curl -L -o inference_config.json https://huggingface.co/wjdghks950/phi1_5_siglip_patch/resolve/main/inference_config.json
549 Bytes
| { | |
| "model_base": "microsoft/phi-1_5", | |
| "base_tokenizer": "microsoft/phi-1_5", | |
| "encoder_type": "siglip", | |
| "attn_level": "patch_layer", | |
| "num_layers_to_choose": 1, | |
| "mm_use_text_enc": true, | |
| "force_reserve_layer": true, | |
| "reserve_layer": -2, | |
| "learnable_gate": false, | |
| "hybrid_reserve_mode": "local_gate", | |
| "hybrid_disagreement_metric": "kl_local_global", | |
| "vision_tower_delta": false, | |
| "conv_mode": "bunny", | |
| "use_router": true, | |
| "temperature": 0, | |
| "num_beams": 1, | |
| "max_new_tokens": 128, | |
| "requires_mol_custom_model_code": true | |
| } | |