google-deepmind/gemma
Google DeepMind's open-weight LLM family with a JAX library for inference, fine-tuning, and multi-modal conversations.

Velocity · 7d
+6.4
★ / day
Trend
→steady
star history
This repository provides the official Python library for Google’s Gemma open-weight language models, built on JAX. It offers APIs for loading checkpoints, running inference with chat samplers, and fine-tuning Gemma models. The library supports multi-turn conversations, multi-modal inputs (images), and multiple Gemma versions from 2B to 27B parameters. It is designed for researchers and developers working with Gemma in JAX-based ML workflows.