README.md

---
base_model: Qwen/Qwen3-4B-Instruct-2507
datasets:
- 'u-10bei/structured_data_with_cot_dataset_512_v2'
- 'u-10bei/structured_data_with_cot_dataset_512_v4'
- 'u-10bei/structured_data_with_cot_dataset_512_v5'
- 'daichira/structured-3k-mix-sft'
- 'daichira/structured-5k-mix-sft'
language:
- en
license: apache-2.0
library_name: peft
pipeline_tag: text-generation
tags:
- qlora
- lora
- structured-output
---

# PiG-v0

This repository provides a **LoRA adapter** fine-tuned from
**Qwen/Qwen3-4B-Instruct-2507** using **QLoRA (4-bit, Unsloth)**.

This repository contains **LoRA adapter weights only**.
The base model must be loaded separately.

## Training Objective

This adapter is trained to improve **structured output accuracy**
(JSON / YAML / XML / TOML / CSV).

Loss is applied only to the final assistant output,
while intermediate reasoning (Chain-of-Thought) is masked.

## Training Configuration

- Base model: Qwen/Qwen3-4B-Instruct-2507
- Method: QLoRA (4-bit)
- Max sequence length: 2048
- Epochs: 1
- Learning rate: 1e-06
- LoRA: r=64, alpha=128

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
import torch

base = "Qwen/Qwen3-4B-Instruct-2507"
adapter = "amu870/PiG-v0"

tokenizer = AutoTokenizer.from_pretrained(base)
model = AutoModelForCausalLM.from_pretrained(
    base,
    torch_dtype=torch.float16,
    device_map="auto",
)
model = PeftModel.from_pretrained(model, adapter)
```

## Sources & Terms (IMPORTANT)

Training data: 
- 'u-10bei/structured_data_with_cot_dataset_512_v2'
- 'u-10bei/structured_data_with_cot_dataset_512_v4'
- 'u-10bei/structured_data_with_cot_dataset_512_v5'
- 'daichira/structured-3k-mix-sft'
- 'daichira/structured-5k-mix-sft'

Dataset License: MIT License and CC-BY-4.0. Visit each dataset repo to check details.