Text Generation
PEFT
Safetensors
English
qwen3
lora
systematic-review
cochrane
title-abstract-screening
medical
conversational
Instructions to use deepcoder2024/Qwen3-4B-LoRA-Cochrane-Screening with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use deepcoder2024/Qwen3-4B-LoRA-Cochrane-Screening with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B") model = PeftModel.from_pretrained(base_model, "deepcoder2024/Qwen3-4B-LoRA-Cochrane-Screening") - Notebooks
- Google Colab
- Kaggle
| { | |
| "model_path": "Qwen/Qwen3-4B", | |
| "train_file": "deepcoder2024/cochrane-screening-sft (train)", | |
| "val_file": "deepcoder2024/cochrane-screening-sft (validation)", | |
| "output_dir": "Qwen3-4B-LoRA-Cochrane-Screening", | |
| "max_length": 2048, | |
| "num_train_epochs": 1.0, | |
| "learning_rate": 0.0002, | |
| "per_device_train_batch_size": 1, | |
| "per_device_eval_batch_size": 1, | |
| "gradient_accumulation_steps": 16, | |
| "warmup_ratio": 0.03, | |
| "weight_decay": 0.01, | |
| "logging_steps": 50, | |
| "eval_steps": 2000, | |
| "save_steps": 2000, | |
| "save_total_limit": 2, | |
| "lora_r": 16, | |
| "lora_alpha": 32, | |
| "lora_dropout": 0.05, | |
| "seed": 42, | |
| "dataloader_num_workers": 2 | |
| } | |