https://github.com/TidalPaladin/combustion

Collection of helpers for deep learning with PyTorch
https://github.com/TidalPaladin/combustion

deep-learning pytorch

Last synced: 2 months ago
JSON representation

Collection of helpers for deep learning with PyTorch

• Host: GitHub
• URL: https://github.com/TidalPaladin/combustion
• Owner: TidalPaladin
• License: apache-2.0
• Created: 2020-04-01T18:01:06.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2022-04-19T13:52:04.000Z (about 3 years ago)
• Last Synced: 2024-10-30T14:26:48.093Z (9 months ago)
• Topics: deep-learning, pytorch
• Language: Python
• Homepage:
• Size: 2.95 MB
• Stars: 4
• Watchers: 2
• Forks: 1
• Open Issues: 6
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Combustion

[![CircleCI](https://circleci.com/gh/TidalPaladin/combustion/tree/master.svg?style=svg)](https://circleci.com/gh/TidalPaladin/combustion/tree/master)

[![Documentation Status](https://readthedocs.org/projects/combustion/badge/?version=latest)](https://combustion.readthedocs.io/en/latest/?badge=latest)

[![codecov](https://codecov.io/gh/TidalPaladin/combustion/branch/master/graph/badge.svg)](https://codecov.io/gh/TidalPaladin/combustion)

Combustion is a collection of layers/helper functions for deep learning.

**NOTE** Highly specialized layers/models are being pruned and will be relocated to separate model-specific repoistories.

See tag `pre-submodule` to return to a state before this pruning.

## Installation

Install combustion core with pip using

```

pip install .

```

Optional dependencies are provided via the following extras:

  * `hdf5` - HDF5 serialization (`combustion.data`)

  * `vision` - Vision helpers (`combustion.vision`)

  * `points` - Point cloud manipulation (`combustion.points`)

  * `macs` - For counting multiply accumulate operations

Development dependencies can be installed with

```

pip install combustion[dev] && pip install combustion[points]

```

## Development

Multiple make recipes are provided to aid in development:

* `ci-test` - Runs tests that will be run by CircleCI

* `ci-quality` - Runs quality checks that will be run by CircleCI

* `docs` - Builds documentation

* `pre-commit` - Installs a pre-commit hook to run code quality tests

* `quality` - Runs code quality tests

* `style` - Automatically formats code (using `black` and `autopep8`)

* `tag-version` - Adds a git tag based on `version.txt`.

* `test` - Runs all tests, including slow model tests

* `test-%` - Runs specific tests by pattern match (via `pytest -k` flag)

* `test-pdb-%` - Runs specific tests with debugging on failure (via `pytest --pdb` flag)

## References

* [PyTorch Lightning](https://github.com/PytorchLightning/pytorch-lightning)