README.md

---
base_model: vectorzhou/gemma-2-2b-it-alpaca-cleaned-SFT
datasets: PKU-Alignment/PKU-SafeRLHF
library_name: transformers
model_name: gemma-2-2b-it-alpaca-cleaned-SFT-PKU-SafeRLHF-Extragradient
tags:
- generated_from_trainer
- text-generation
- fine-tuned
- trl
- extra-gradient
licence: license
---

# Model Card for gemma-2-2b-it-alpaca-cleaned-SFT-PKU-SafeRLHF-Extragradient

This model is a fine-tuned version of [vectorzhou/gemma-2-2b-it-alpaca-cleaned-SFT](https://huggingface.co/vectorzhou/gemma-2-2b-it-alpaca-cleaned-SFT) on the [PKU-Alignment/PKU-SafeRLHF](https://huggingface.co/datasets/PKU-Alignment/PKU-SafeRLHF) dataset.
It has been trained using [TRL](https://github.com/huggingface/trl).

## Quick start

```python
from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="vectorzhou/gemma-2-2b-it-alpaca-cleaned-SFT-PKU-SafeRLHF-Extragradient-0420125020-epoch-1", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])
```

## Training procedure

[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/zhourunlongvector/nlhf/runs/855dswk0) 


This model was trained with Extragradient, a method introduced in [Extragradient Preference Optimization (EGPO): Beyond Last-Iterate Convergence for Nash Learning from Human Feedback](https://huggingface.co/papers/2503.08942).

### Framework versions

- TRL: 0.13.0
- Transformers: 4.48.0
- Pytorch: 2.2.1
- Datasets: 3.2.0
- Tokenizers: 0.21.0

## Citations

Cite Extragradient as:

```bibtex
@misc{zhou2025extragradientpreferenceoptimizationegpo,
    title={Extragradient Preference Optimization (EGPO): Beyond Last-Iterate Convergence for Nash Learning from Human Feedback}, 
    author={Runlong Zhou and Maryam Fazel and Simon S. Du},
    year={2025},
    eprint={2503.08942},
    archivePrefix={arXiv},
    primaryClass={cs.LG},
    url={https://arxiv.org/abs/2503.08942}, 
}
```

Cite TRL as:
    
```bibtex
@misc{vonwerra2022trl,
	title        = {{TRL: Transformer Reinforcement Learning}},
	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
	year         = 2020,
	journal      = {GitHub repository},
	publisher    = {GitHub},
	howpublished = {\url{https://github.com/huggingface/trl}}
}
```