CFGauss/unsloth_17-150

unsloth_qwen3-4b-structured-output-lora-unsloth-17

This repository provides a LoRA adapter fine-tuned from unsloth/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: unsloth/Qwen3-4B-Instruct-2507
• Method: QLoRA (4-bit)
• Max sequence length: 2048
• Epochs: 2
• Learning rate: 1e-04
• LoRA: r=128, alpha=256

Usage

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

base = "unsloth/Qwen3-4B-Instruct-2507"
adapter = "your_id/your-repo"

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_v5

• daichira/structured-5k-mix-sft
• daichira/structured-hard-sft-4k

Dataset License: MIT License and CC BY 4.0.

u-10bei/structured_data_with_cot_dataset_512_v2 is licensed under the MIT License.

daichira/structured-5k-mix-sft and daichira/structured-hard-sft-4k is licensed under the CC BY 4.0 License.