Releases: NLeSC/mcfly
v4.1.0
What's Changed
- #312 - adding sphinx build action by @lyashevska in #317
- Fix read the doc build by @maltelueken in #319
- Change CITATION.cff cff-version to 1.2.0 by @abelsiqueira in #318
- refactor modelgen by @cwmeijer in #304
- Add multi-backend Keras 3 support by @HeinrichAD in #325
New Contributors
- @abelsiqueira made their first contribution in #318
- @HeinrichAD made their first contribution in #325
Full Changelog: v4.0.0...v4.1.0
v4.0.0
v3.1.0
- Added support for training with Datasets, generators, and other data types supported by Keras #211
- Added separate classes for each model type, also allowing easier extension by own custom models #239
- Dropped support for Python 3.5 (supported are: Python 3.6, 3.7 and 3.8)
- Extend and update documentation in readthedocs #250
- Fix broken model visualization #256
v3.0.0
mcfly 3.0 lets you train new network types on your time-series data. It brings support for ResNets and InceptionTime architectures next to the already supported CNN and DeepConvLSTM architectures. Version 3.0 also brings API improvements, bugfixes and uses Tensorflow 2.0 under the hood.
v2.1.0
v2.0.1
This release makes the switch from working on top of Keras (with Tensorflow or Theano underneath) to working on top of Tensorflow directly. Because of this dependency change saved models with earlier versions are no longer compatible with the current version.
v2.0.0
This release makes the switch from working on top of Keras (with Tensorflow or Theano underneath) to working on top of Tensorflow directly. Because of this dependency change saved models with earlier versions are no longer compatible with the current version.
v1.0.5
v1.0.4
Improved packaging, now citable
This release is a small patch over v1.0.2. It fixes configuration with Zenodo so the release gets assigned a DOI. Also broadened the support for versions of dependency: six.