#175 · Primary category: MLOps & Evaluation

talos

artificial-intelligence deep-learning hyperparameter-optimization keras keras-tensorflow tensorflow

Hyperparameter Experiments with TensorFlow and Keras

Project last updated:04/22/24

GitHub Stars

1.6K

Forks

266

Contributors

23

License

MIT

Why we included this project

Talos targets the everyday TensorFlow and Keras workflow where you keep editing a script to try another learning rate or layer size, then rerun it by hand. Wrap your existing model definition in a scan call, hand it a grid of parameter values, and Talos runs the sweeps and reports which configurations performed best. It adds no new syntax and works with any standard Keras, tf.keras, or PyTorch model, so the code stays yours and you remain in control. For people who want tidy hyperparameter studies without standing up a heavyweight orchestration platform, it is a small, pragmatic tool.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category