A framework for code-agnostic, interactive prototyping of DNNs.

Features

• Transparent and elastic scheduling of DNN training jobs on modern HPC systems.
• Monitoring and visualizing model parameters and computational performance statistics.
• Perform semi-automatic hyperparameter tuning/optimization and architecture search using evolutionary algorithms.
• A user-defined interactive interface to drive the framework/ design process, not bound to any particular framework.
• Scaling the functionality and performance of the model as the resources increase.

How do I get set up?

• pip3 install protonn for latest stable release
• pip3 install git+https://github.com/protoNN-ai/protoNN.git for recent development version
• Python 3.6 or later is required

Contributors

Aleksandr Drozd

Mohamed Wahib

Mateusz Bysiek

Maxim Shpakovich

For licensing information, please see LICENSE