gage

Model evaluation and publishing toolkit


License
Apache-2.0
Install
pip install gage==0.1.2

Documentation

Gage

Gage is a tool for running model evaluations and publishing results. It's under active development by the team that developed Guild AI.

Installing

Install the latest version of Gage using pip:

pip install gage

Contributing

We warmly encourage contributions to this project at any level. We are committed to growing and supporting a rich, diverse set of contributors. If you would like to contribute something, feel free to open an issue to ask questions, make suggestions, or otherwise let us know what you're thinking. Changes to the project are accepted using GitHub pull requests.

Please read Contributor Covenant Code of Conduct to familiarize yourself with the project code of conduct and contribution policy. If you feel that something is missing from this document or could be improved, please let us know by either opening an issue or by emailing the project administrator at garrett@placeon.earth. Private correspondence via email will be held in strict confidence.

Setting up a Local Development Environment

  1. Fork the source code repository at https://github.com/gageml/gage

  2. Create a project-specific virtual environment. Use Python 3.10 or later.

    cd gage
    virtualenv --python python3.10 .venv  # Any version >= 3.10 is okay
  3. Activate the virtual environment and use pip to install the project as "editable".

    source .venv/bin/activate  # Works on most POSIX shells - change as needed
    pip install -e '.[dev]'    # Installs gage and its dev requirements
  4. Run tests using Groktest (optional).

    groktest .