---
base_model: motobrew/qwen3-adv-comp-v34
datasets:
- motobrew/alf-dpo-from-top-alf93-v0
language:
- en
license: apache-2.0
library_name: transformers
pipeline_tag: text-generation
tags:
- dpo
- unsloth
- qwen
- alignment
---

# qwen-dpo-v3

This model is a fine-tuned version of **motobrew/qwen3-adv-comp-v34** using **Direct Preference Optimization (DPO)** via the **Unsloth** library.

## Training Objective
This model has been optimized using DPO to align its responses with preferred outputs, focusing on improving reasoning (Chain-of-Thought) and structured response quality based on the provided preference dataset.

## Training Configuration
- **Base model**: motobrew/qwen3-adv-comp-v34
- **Method**: DPO (Direct Preference Optimization)
- **Epochs**: 1
- **Learning rate**: 2e-06
- **Beta**: 0.02
- **Max sequence length**: 1024

## Usage
You can use this model directly with `transformers`.

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

model_id = "motobrew/qwen-dpo-v3"

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**: [motobrew/alf-dpo-from-top-alf93-v0]
* **License**: MIT License. (As per dataset terms).
* **Compliance**: Users must follow the original base model's license terms.