TRL

Train transformer language models with reinforcement learning: SFT, DPO, PPO, and GRPO.

Stars

19,688

Forks

3,124

Language

Python

Get the code

# Clone the repository
git clone https://github.com/huggingface/trl

What is TRL?

Train transformer language models with reinforcement learning: SFT, DPO, PPO, and GRPO.

Is TRL safe to use?

TRL is an open-source project by huggingface on GitHub. Review the source code and license before using it in production.

How do I get started with TRL?

git clone https://github.com/huggingface/trl

What language is TRL written in?

openSource.faqWhatLanguageAnswer

Fine-tune and run LLMs and diffusion models 2x faster with far less memory.

Unified efficient fine-tuning of 100+ LLMs and VLMs with a no-code web UI.

Deep learning optimization library that makes distributed training and inference easy and efficient.