google-gemma/cookbook

A collection of guides and examples for the Gemma open models from Google.

What it solves

The Gemma Cookbook provides a centralized collection of practical guides, tutorials, and code examples to help developers build applications using Google's Gemma family of open AI models. It addresses the challenge of learning how to implement, fine-tune, and deploy these models effectively.

How it works

The repository is organized into thematic folders containing Jupyter notebooks and technical documentation. It covers everything from core model usage and full-stack application demos to research-focused experiments and responsible AI development practices. Developers can use these resources to understand how to work with various Gemma variants, including those specialized for coding, medical analysis, safety, and multimodal tasks.

Who it’s for

This project is for developers and researchers who want to build with Google's open Gemma models. Whether you are a beginner looking for a starting point or an advanced user exploring specialized variants like MedGemma or ShieldGemma, the cookbook offers tested code to accelerate your development.

Highlights

  • Comprehensive tutorials for core Gemma models and specialized variants.
  • Full-stack application demos for end-to-end use cases.
  • Research-focused notebooks covering experimental models like TxGemma.
  • Dedicated resources for responsible and safe AI development.
  • Technical documentation on model capabilities and integration guides.

Related

  • Project
    IDEA-CCNL/Fengshenbang-LM

    An open-source ecosystem of Chinese pre-trained foundation models and a supporting framework for natural language understanding, generation, and multimodal tasks.

  • Project
    ombharatiya/ai-system-design-guide

    A continuously updated, MIT-licensed reference for production AI system design and AI engineering interviews — covering RAG, agents, model selection, evaluation, security, and 20 real-world case studies, plus a 147-question interview bank.

  • Project
    groupultra/telegram-search

    A personal Telegram message backup and search tool that uses vector embeddings and RAG to enable semantic search and AI-driven summaries of chat histories.

  • Project
    marcelroed/gigatoken

    A high-performance Rust-based tokenizer for language modeling that provides up to 1000x speedup over HuggingFace tokenizers by using SIMD-optimized pretokenization.