Tagger decoder config override and auxiliary validation checks #3222
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
for more information, see https://pre-commit.ci
justinxzhao
approved these changes
Mar 8, 2023
abidwael
approved these changes
Mar 8, 2023
w4nderlust
approved these changes
Mar 9, 2023
justinxzhao
pushed a commit
that referenced
this pull request
Mar 22, 2023
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
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Outdated suggestions cannot be applied.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.
This PR ensures that tagger decoders for text and sequence features are used correctly within Ludwig. Here are the the main changes addressed by this PR:
reduce_output=None, then it will raise a ConfigValidationError because the tagger decoder won't receive a b x s x h (3D) input from the combiner.reduce_inputfor text/sequence output features to be set to None to prevent the hidden tensors from being reshaped to 2D. The default value forreduce_inputis set tosumsince it is the more popular case, but will be overridden during ModelConfig object creation when a tagger decoder is being used, and logs a warning. This is because this is an implementation detail Ludwig users should not have to worry about since it relates to tensor shapes directly.I've also added a few different tests for text and sequence output features, as well as a test for a config that does not have a text/sequence/timeseries input feature to make sure all of these different cases are correctly tested and accounted for.