Matsuo2025-Main-Qwen3-4B-DPO-v1

Matsuo2025-Main-Qwen3-4B-DPO-v1

This model is a fine-tuned version of Matsuo2025-Main-Qwen3-4B-SFT-v1 using Direct Preference Optimization (DPO). The LoRA weights have been merged into the base model (16-bit).

Training Configuration

• Base model: RegFo/Matsuo2025-Main-Qwen3-4B-SFT-v1
• Method: DPO (Direct Preference Optimization)
• Dataset: u-10bei/dpo-dataset-qwen-cot

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "RegFo/Matsuo2025-Main-Qwen3-4B-DPO-v1"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")

# Test inference
prompt = "Your question here"
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.