Gemma Multimodal Fine-Tuner

Fine-tune Hugging Face Gemma models on text, images, and audio using PEFT LoRA. The repository includes a command-line wizard and training tools for Apple Silicon.

Install

python3 -m venv .venv
source .venv/bin/activate
pip install -e .
cp config/config.ini.example config/config.ini

Review the configuration before running training. Model downloads may require Hugging Face authentication and acceptance of the model license.

Use

gemma-macos-tuner --help
gemma-macos-tuner system-check

Implementation lives in gemma_tuner/; additional tools live in tools/. See pyproject.toml for dependencies and command-line entry points.

Repository scope

This public repository contains code and public usage documentation. Personal plans, research notes, experiment receipts, and agent configuration are maintained separately and are excluded here. Contributions must use clean code changes; do not merge branches containing private development history.

License

See LICENSE. Bundled third-party code retains its own license notices.