rnnr

rnnr: neural network runner

rnnr helps you to run your neural network models, for either training or evaluation. It is heavily inspired by Ignite and torchnet, but hopefully simpler and more applicable to libraries other than PyTorch.

Documentation

https://rnnr.readthedocs.io

Contributing

Pull requests are welcome! To start contributing, first install flit.

pip install flit

Next, install this library and its dependencies in development mode.

flit install --symlink

Lastly, setup the pre-commit hook.

ln -s ../../pre-commit.sh .git/hooks/pre-commit

Tests and the linter can be run with pytest and flake8 respectively. The latter also runs mypy for type checking.

License

Apache License, Version 2.0