Welcome

See our library Tributaries for mass-deploying UnifiedML apps on remote servers.

Check out minihydra / leviathan for how we handle sys args & hyperparams.

Install

pip install UnifiedML

What is UnifiedML?

UnifiedML is a toolbox & engine for defining ML tasks and training them individually, or together in a single general intelligence.

See full documentation here. (Documentation in progress)

Get started

To start a train session, use the ML command:

ML

Defaults:

dataset=MNIST

task=classify

Plots, logs, generated images, and videos are automatically stored in: ./Benchmarking.

Voluntary support

Sponsoring

This library is built by a single individual from low-income means on the funding of a graduate student wage. Please support financially by sponsoring if you can and if you have access to compute and resources that you would be willing to contribute, please reach out to me at slerman12@gmail.com.

Contributing code

This library is built with passion and dedication, and every effort has been made to make it minimal and elegant. The standards for accepted push requests are high, but if you have a push that you would like to make or a contribution that you would like to propose, you may make a request and/or leave a note to me either in issues or by email at slerman12@gmail.com.

Thank you.