luz 9.0.0

Framework for rapid research and development of machine learning projects using PyTorch.

Framework for rapid research and development of machine learning projects using PyTorch.

Longer description coming soon!

Important features:

1. Reduced boilerplate and simple specifications of complex workflows

2. Easy and flexible customization of Learner by overriding methods

3. Built-in scoring algorithms like holdout and cross validation

4. Straightforward hyperparameter tuning with built-in tuning algorithms like random search and grid search

5. Model.

6. Training scheme.

7. Overall learning algorithm.

1. Hyperparameter selection.

1. Unified development interface through Learner object. Simply inherit luz.Learner, define the model, loader, and param functions, and you’re good to go. Add a hyperparams function to enable tuning and make tuned parameters accessible in the model and param functions.

Getting Started

Prerequisites

Installing

To install, open a shell terminal and run:

`conda create -n luz -c conda-forge -c pytorch -c kijana luz`

Versioning

Authors

Ki-Jana Carter

License

This project is licensed under the MIT License - see the LICENSE file for details.