Unsloth now supports full fine-tuning, 8-bit and all models! 🦥

Unsloth Documentation

Homepage Reddit Discord

Llama 4 Gemma 3 GRPO (Reasoning)Dynamic 4-bit Vision fine-tuning

Homepage Reddit Discord

Blog Llama 4 Gemma 3 GRPO (Reasoning)Dynamic 4-bit Vision fine-tuning Sign up

Get Started
Basics

Powered by GitBook

On this page

Was this helpful?

Get Started

Installing + Updating

Learn to install Unsloth locally or online.

PreviousAll Our Models NextUpdating

Last updated 21 days ago

Was this helpful?

Unsloth works on Linux, Windows directly, Kaggle, Google Colab and more. See our .

Recommended installation method:

pip install unsloth

📥

system requirements

Windows Installation