Skip to content
Lightweight framework for fast prototyping and training deep neural networks with PyTorch and TensorFlow
Python Jupyter Notebook Other
Branch: master
Clone or download
Pull request Compare This branch is 4 commits ahead, 204 commits behind delira-dev:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github Update unittests.yml Sep 11, 2019
delira PEP-8 Auto-Fix Sep 10, 2019
docker Update Dockerfile Jan 15, 2019
docs Merge branch 'master' into slack Aug 29, 2019
notebooks Adjust comments Aug 7, 2019
paper reformat references Jun 6, 2019
requirements Merge master into remove_trixi2 Sep 9, 2019
scripts/ci Fix slack version to 1.3.1 Aug 29, 2019
tests PEP-8 Auto-Fix Sep 10, 2019
.codecov.yml Update .codecov.yml Jan 22, 2019
.gitattributes Add versioneer Jul 9, 2019
.gitignore Fix metric test Jun 6, 2019
.readthedocs.yml Tf integration (delira-dev#51) Feb 6, 2019
.travis.yml Move coverage to tests Sep 6, 2019
AUTHORS.rst Add @mibaumgartner to core developers Jun 1, 2019
CODEOWNERS Update CODEOWNERS Jul 31, 2019
CONTRIBUTING.md
LICENSE
MANIFEST.in Merge branch 'master' into additional_backends Jul 29, 2019
README.md Change string refering to available backends Aug 5, 2019
pytest.ini enable tf tests Feb 21, 2019
setup.cfg Changes suggestions Sep 10, 2019
setup.py Merge branch 'master' into additional_backends Jul 29, 2019
versioneer.py

README.md

PyPI version Build Status Documentation Status codecov LICENSE DOI

logo

delira - A Backend Agnostic High Level Deep Learning Library

Authors: Justus Schock, Michael Baumgartner, Oliver Rippel, Christoph Haarburger

Introduction

delira is designed to work as a backend agnostic high level deep learning library. You can choose among several computation backends. It allows you to compare different models written for different backends without rewriting them.

For this case, delira couples the entire training and prediction logic in backend-agnostic modules to achieve identical behavior for training in all backends.

delira is designed in a very modular way so that almost everything is easily exchangeable or customizable.

A (non-comprehensive) list of the features included in delira:

  • Dataset loading
  • Dataset sampling
  • Augmentation (multi-threaded) including 3D images with any number of channels (based on batchgenerators)
  • A generic trainer class that implements the training process for all backends
  • Training monitoring using Visdom or Tensorboard
  • Model save and load functions
  • Already impelemented Datasets
  • Many operations and utilities for medical imaging

What about the name?

delira started as a library to enable deep learning research and fast prototyping in medical imaging (especially in radiology). That's also where the name comes from: delira was an acronym for DEep Learning In RAdiology*. To adapt many other use cases we changed the framework's focus quite a bit, although we are still having many medical-related utilities and are working on constantly factoring them out.

Installation

Choose Backend

You may choose a backend from the list below. If your desired backend is not listed and you want to add it, please open an issue (it should not be hard at all) and we will guide you during the process of doing so.

Backend Binary Installation Source Installation Notes
None pip install delira pip install git+https://github.com/delira-dev/delira.git Training not possible if backend is not installed separately
torch pip install delira[torch] git clone https://github.com/delira-dev/delira.git && cd delira && pip install .[torch] delira with torch backend supports mixed-precision training via NVIDIA/apex (must be installed separately).
torchscript pip install delira[torchscript] git clone https://github.com/delira-dev/delira.git && cd delira && pip install .[torchscript] The torchscript backend currently supports only single-GPU-training
tensorflow eager pip install delira[tensorflow] git clone https://github.com/delira-dev/delira.git && cd delira && pip install .[tensorflow] the tensorflow backend is still very experimental and lacks some features
tensorflow graph pip install delira[tensorflow] git clone https://github.com/delira-dev/delira.git && cd delira && pip install .[tensorflow] the tensorflow backend is still very experimental and lacks some features
scikit-learn pip install delira pip install git+https://github.com/delira-dev/delira.git /
chainer pip install delira[chainer] git clone https://github.com/delira-dev/delira.git && cd delira && pip install .[chainer] /
Full pip install delira[full] git clone https://github.com/delira-dev/delira.git && cd delira && pip install .[full] All backends will be installed.

Docker

The easiest way to use delira is via docker (with the nvidia-runtime for GPU-support) and using the Dockerfile or the prebuild-images.

Chat

We have a community chat on slack. If you need an invitation, just follow this link.

Getting Started

The best way to learn how to use is to have a look at the tutorial notebook. Example implementations for classification problems, segmentation approaches and GANs are also provided in the notebooks folder.

Documentation

The docs are hosted on ReadTheDocs/Delira. The documentation of the latest master branch can always be found at the project's github page.

Contributing

If you find a bug or have an idea for an improvement, please have a look at our contribution guideline.

You can’t perform that action at this time.