Add XTransformer backend #716
Draft
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Kudos, SonarCloud Quality Gate passed!
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The code in this PR is by @mo-fu and originally submitted via PR #540. That PR got accidentally closed and could not be re-opened, which is why this new PR needs to be opened for the XTransformer backend. (This PR is coming from the point of the git history just before the unsuccessful commits attempting to make the original PR re-openable.)
The description of the original PR is below.
This PR adds XTransformer as an optional backend to Annif. For now it does not yet use distilbert in the default configuration as this is not yet available on pypi.
The tests for the backend resort to mocking as training would download a pretrained model of size at least 500 mb.
Also we should discuss cache directories. At the moment xtransformer will download models from the huggingface hub to
~/.cache/huggingfaceIs this behavior desired for Annif or should the cache be placed in the data folder?I also haven't modified the docker container yet. When I installed pecos in a venv it required BLAS libraries so this would probably have to be added to the container. Additionally pecos will install the GPU enabled pytorch. Meaning the container size will grow. Therefore I wanted to check with you first before adding it.