add: metacat can predict on spans in arbitrary spangroups #391
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
jkgenser
changed the title
mart-r
requested changes
Jan 23, 2024
Thank you for your comments! I will address them. |
|
@mart-r I have iterated on your PR comments. I also added a more informative warning in case the span group is not set no the doc and a test for it. |
I've done my best to address all of your changes. Let me know if you have further comments! |
mart-r
approved these changes
Jan 29, 2024
@mart-r Got it, I fixed the linting and flake8 passes locally. |
mart-r
approved these changes
Jan 30, 2024
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Outdated suggestions cannot be applied.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.
Hello - I opened this PR request to start a discussion about adding a new feature that allows a trained MetaCAT model to run on arbitrary spangroup. This will will allow configuring a MetaCAT model such that it can be pointed at an arbitrary spacy span group and set meta_anns on it.
Currently, it is allowed to run on the two set of
list[spacy.token.Span]:doc.entsordoc._.ents.However, there are use cases where we may store other spans as part of a pipeline under
doc.spans["my_spangroup_name"]that are distinct from doc.ents in which we want to call predict on the meta cat model.I don't think it's yet necessary to be able to train on arbitrary spangroups since that would get into changing the input format or extract format from medcat trainer.
One workaround is to either zero out the ents object and set the spangroup to those ents and call predict however this starts to feel hacky.
Another workaround is to create a duplicate
spacy.token.Docobject and usingdoc.entsfor the separate spangroup. However this is inefficient as it requires running parts of the spacy pipeline again and means we can't use this spangroup in other pipeline components.