Text Generation
PEFT
Safetensors
English
named-entity-recognition
chain-of-thought
lora
conversational
Instructions to use structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B-Thinking-2507") model = PeftModel.from_pretrained(base_model, "structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora") - Notebooks
- Google Colab
- Kaggle
Download training_args.json from structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora: direct link, hf CLI and curl.
- Browser
- Download file 512 Bytes
-
https://huggingface.co/structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora/resolve/main/training_args.json
- Command line
-
hf download hf://structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora/training_args.json
-
curl -L -o training_args.json https://huggingface.co/structured-cot-ner/qwen3-4b-thinking-structured-noedge-cot-lora/resolve/main/training_args.json
512 Bytes
| { | |
| "data_path": "/root/code/tutor_data/noedge.jsonl", | |
| "model_id": "Qwen/Qwen3-4B-Thinking-2507", | |
| "output_dir": "/outputs/work/qwen3-4b-thinking_noedge", | |
| "load_in_4bit": false, | |
| "n_train_samples": null, | |
| "validation_size": 0.05, | |
| "max_seq_length": 5120, | |
| "num_epochs": 2, | |
| "lora_r": 16, | |
| "lora_alpha": 32, | |
| "learning_rate": 0.0001, | |
| "per_device_batch": 1, | |
| "grad_accum_steps": 8, | |
| "eval_steps": 100, | |
| "seed": 42, | |
| "hf_repo_id": null, | |
| "skip_push": true, | |
| "max_steps": -1, | |
| "use_cpu": false | |
| } |