PEFT
Safetensors
English
reasoning
mathematical-reasoning
reward-model
preference-model
prefix-evaluation
utility-model
lora
qwen
Instructions to use zhiqix/PUM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use zhiqix/PUM with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
| { | |
| "model_name_or_path": "Qwen/Qwen3-4B-Instruct-2507", | |
| "trust_remote_code": true, | |
| "dropout": 0.1, | |
| "train_mode": "lora_head", | |
| "lora_r": 64, | |
| "lora_alpha": 128, | |
| "lora_dropout": 0.1, | |
| "lora_target_modules": null, | |
| "value_head_type": "mlp" | |
| } |