Select valid BERT QA answer spans (#684)#1852
Open
craymichael wants to merge 1 commit into
Open
Conversation
Issue: meta-pytorch#684 Issue summary: The BERT SQuAD tutorials selected the start and end tokens independently. If the highest-scoring end token came before the highest-scoring start token, the printed answer span was empty. Context: In extractive question answering, a span is the contiguous slice of input tokens used as the answer. The start token is the first token in that slice, and the end token is the last token. A valid answer must satisfy end_idx >= start_idx. Summary: - Select the best valid answer span by scoring every start/end token pair and masking pairs where the end token precedes the start token. - Apply the same span selection to both BERT SQuAD tutorial notebooks. - Add a notebook regression test for invalid independent argmax spans. Test Plan: - venv/uv/bin/python -m pytest -q tests/test_tutorial_regressions.py - Pyre: not applicable; notebook-only behavior change covered by regression test.
Contributor
|
@craymichael has imported this pull request. If you are a Meta employee, you can view this in D106053849. |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes #684.
Issue: #684
Issue summary: The BERT SQuAD tutorials selected the start and end tokens independently. If the highest-scoring end token came before the highest-scoring start token, the printed answer span was empty.
Context:
In extractive question answering, a span is the contiguous slice of input tokens used as the answer. The start token is the first token in that slice, and the end token is the last token. A valid answer must satisfy end_idx >= start_idx.
Summary:
Test Plan: