This model is a LoRA adapter fine-tuned from Qwen/Qwen3-4B-Instruct-2507 using Direct Preference Optimization (DPO) via the Unsloth library.

Training Objective

This adapter aligns responses with preferred outputs from the provided preference dataset.

Training Configuration

Base model: Qwen/Qwen3-4B-Instruct-2507
Method: DPO (Direct Preference Optimization)
Epochs: 1
Learning rate: 2e-06
Beta: 0.1
Max sequence length: 768
LoRA Config: r=16, alpha=32

Usage (Adapter)

python

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

base_id = "Qwen/Qwen3-4B-Instruct-2507"
adapter_id = "your_id/your-adapter-repo"

model = AutoModelForCausalLM.from_pretrained(
    base_id,
    torch_dtype=torch.float16,
    device_map="auto",
    trust_remote_code=True,
)
model = PeftModel.from_pretrained(model, adapter_id)

# Test inference
prompt = "Your question here"
tokenizer = AutoTokenizer.from_pretrained(base_id, trust_remote_code=True)
inputs = tokenizer.apply_chat_template([{"role": "user", "content": prompt}], tokenize=True, add_generation_prompt=True, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0]))

Sources & License (IMPORTANT)

Training Data: [u-10bei/dpo-dataset-qwen-cot]
License: MIT License. (As per dataset terms).
Compliance: Users must follow the original base model's license terms.

yonetaro/qwen3-4b-structured-lora-v2-merged

qwen3-4b-structured-output-lora-yk

Training Objective

Training Configuration

Usage (Adapter)

Sources & License (IMPORTANT)

yonetaro/
qwen3-4b-structured-lora-v2-merged