https://github.com/google-research/kauldron

Modular, scalable library to train ML models
https://github.com/google-research/kauldron

deep-learning jax python

Last synced: about 1 year ago
JSON representation

Modular, scalable library to train ML models

• Host: GitHub
• URL: https://github.com/google-research/kauldron
• Owner: google-research
• License: apache-2.0
• Created: 2023-06-19T08:56:33.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2025-04-09T17:56:22.000Z (about 1 year ago)
• Last Synced: 2025-04-09T18:53:01.729Z (about 1 year ago)
• Topics: deep-learning, jax, python
• Language: Python
• Homepage: https://kauldron.readthedocs.io/
• Size: 1.78 MB
• Stars: 99
• Watchers: 7
• Forks: 13
• Open Issues: 51
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# kauldron

[![Unittests](https://github.com/google-research/kauldron/actions/workflows/pytest_and_autopublish.yml/badge.svg)](https://github.com/google-research/kauldron/actions/workflows/pytest_and_autopublish.yml)

[![PyPI version](https://badge.fury.io/py/kauldron.svg)](https://badge.fury.io/py/kauldron)

[![Documentation Status](https://readthedocs.org/projects/kauldron/badge/?version=latest)](https://kauldron.readthedocs.io/en/latest/?badge=latest)

Kauldron is a library for training machine learning models, optimized for

**research velocity** and **modularity**.

**Modularity**:

*   All parts of Kauldron are self-contained, so can be used independently

    outside Kauldron.

*   Use any dataset (TFDS, Grain, SeqIO, your custom pipeline),

    any (flax) model, any optimizer,... Kauldron provides the

    glue that link everything together.

*   Everything can be customized and overwritten (e.g. sweep over models

    architecture, overwrite any inner layer parameter,...)

**Research velocity**:

*   Everything should work out-of the box. The

    [example configs](https://github.com/google-research/kauldron/tree/main/examples/mnist_autoencoder.py)

    can be used and customized as a starting point.

*   Colab-first workflow for easy prototyping and fast iteration

*   Polished user experience (integrated XM plots, profiler,

    post-mortem debugging on borg, runtime shape checking, and many others...).

[Open an issue](https://github.com/google-research/kauldron/issues)..

*This is not an officially supported Google product.*