Preparation for release of skorch v0.12.0

[BenjaminBossan]

Changes:

Bump version

Update CHANGES.md

Remove warning about signature change in on_batch methods

We had a change in the signature of on_batch_begin and on_batch_end.
Therefore, we tried to detect if users had overridden these methods and
issued a (hopefully) helpful warning.

This change is now established long enough that we can remove the
warning and associated tests.

Release text

@ottonemo and I are working on it, will paste it once it's ready.

[BenjaminBossan]

proposed release text:

We're pleased to announce a new skorch release, bringing new features that might interest you.

The main changes relate to better integration with the Hugging Face ecosystem:

• Benefit from faster training and inference times thanks to easy integration with accelerate via skorch's AccelerateMixin.
• Better integration of tokenizers via skorch's HuggingfaceTokenizer and HuggingfacePretrainedTokenizer; you can even put Hugging Face tokenizers into an sklearn Pipeline and perform a grid search to find the best tokenizer hyperparameters.
• Automatically upload model checkpoints to Hugging Face Hub via skorch's HfHubStorage.
• Check out this notebook to see how to use skorch and Hugging Face together.

But this is not all. We have added the possibility to load the best model parameters at the end of training when using the EarlyStopping callback. We also added the possibility to remove unneeded attributes from the net after training when it is intended to be only used for prediction by calling the trim_for_prediction method. Moreover, we now show how to use skorch with PyTorch Geometric in this notebook.

As always, this release was made possible by outside contributors. Many thanks to:

• Alan deLevie (@adelevie)
• Cédric Rommel (@cedricrommel)
• Florian Pinault (@floriankrb)
• @terminator-ger
• Timo Kaufmann (@timokau)
• @TrellixVulnTeam

Find below the list of all changes:

Added

• Added load_best attribute to EarlyStopping callback to automatically load module weights of the best result at the end of training
• Added a method, trim_for_prediction, on the net classes, which trims the net from everything not required for using it for prediction; call this after fitting to reduce the size of the net
• Added experimental support for huggingface accelerate; use the provided mixin class to add advanced training capabilities provided by the accelerate library to skorch
• Add integration for Huggingface tokenizers; use skorch.hf.HuggingfaceTokenizer to train a Huggingface tokenizer on your custom data; use skorch.hf.HuggingfacePretrainedTokenizer to load a pre-trained Huggingface tokenizer
• Added support for creating model checkpoints on Hugging Face Hub using HfHubStorage
• Added a notebook that shows how to use skorch with PyTorch Geometric (torch geometric support + example notebook #863)

Changed

• The minimum required scikit-learn version has been bumped to 0.22.0
• Initialize data loaders for training and validation dataset once per fit call instead of once per epoch (migration guide)
• It is now possible to call np.asarray with SliceDatasets (Bugfix: sklearn 1.1 and SliceDataset #858)

Fixed

• Fix a bug in SliceDataset that prevented it to be used with to_numpy (Bugfix: sklearn 1.1 and SliceDataset #858)
• Fix a bug that occurred when loading a net that has device set to None (Nets now can be loaded even if device=None #876)
• Fix a bug that in some cases could prevent loading a net that was trained with CUDA without CUDA
• Enable skorch to work on M1/M2 Apple MacBooks (Move mps data to CPU in util.to_numpy #884)

[thomasjpfan]

LGTM

As for the release notes, there is no migration guide in the FAQ for going from 0.11 to 0.12. Specifically the following does not link to a section: https://skorch.readthedocs.io/en/stable/user/FAQ.html#migration-from-0-11-to-0-12

[BenjaminBossan]

As for the release notes, there is no migration guide in the FAQ for going from 0.11 to 0.12. Specifically the following does not link to a section: https://skorch.readthedocs.io/en/stable/user/FAQ.html#migration-from-0-11-to-0-12

Yes, but as soon as we release, the stable docs should be updated and the link should work.