TRL

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

Sterren

19,688

Forks

3,124

Taal

Python

Categorie

Training & Fine-tuning

Code ophalen

# Kloon de repository
git clone https://github.com/huggingface/trl

Wat is TRL?

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

Is het gebruik van TRL veilig?

TRL is een open-sourceproject van huggingface op GitHub. Bekijk de broncode en licentie voordat je het in productie gebruikt.

Hoe begin ik met TRL?

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

In welke taal is TRL geschreven?

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.