feat: added the precision and recall metrics for QA accuracy #157
Conversation
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
franluca
reviewed
Dec 14, 2023
| """ | ||
| if normalize_text: # pragma: no branch | ||
| model_output, target_output = (_normalize_text_quac_protocol(text) for text in (model_output, target_output)) | ||
| ret = precision(reference=set(target_output.split(" ")), test=set(model_output.split(" "))) |
There was a problem hiding this comment.
why set and not list? we want to discard repetitions of words?
There was a problem hiding this comment.
Valid question, but set seems standard. At least that's what the NLTK metric assumes that we use here.
| """ | ||
| if normalize_text: # pragma: no branch | ||
| model_output, target_output = (_normalize_text_quac_protocol(text) for text in (model_output, target_output)) | ||
| ret = recall(reference=set(target_output.split(" ")), test=set(model_output.split(" "))) |
polaschwoebel
previously approved these changes
Dec 14, 2023
| """ | ||
| if normalize_text: # pragma: no branch | ||
| model_output, target_output = (_normalize_text_quac_protocol(text) for text in (model_output, target_output)) | ||
| ret = precision(reference=set(target_output.split(" ")), test=set(model_output.split(" "))) |
There was a problem hiding this comment.
Valid question, but set seems standard. At least that's what the NLTK metric assumes that we use here.
polaschwoebel
previously approved these changes
Dec 15, 2023
franluca
previously approved these changes
Dec 15, 2023
malhotra18
previously approved these changes
Dec 15, 2023
d36ef85
malhotra18
approved these changes
Jan 11, 2024
polaschwoebel
approved these changes
Jan 12, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
Description of changes: This pull request adds the Precision and Recall metrics for the Question Answering task. Previous metrics do not capture the cases where one of the target output or model output is short and the other one is long.
For instance, consider the question
Did RMS Titanic sink in 1912?If the target output isYesand the model output isYes. The ship indeed sank in 1912. It was the largest ship at the time <some long text>then the existing metrics will give a low score even though the answer is correct. The recall metric added in this PR will be1.0indicating that all of the target output words are contained within the model output. The precision metric operates in the opposite direction and measures what fraction of words in the model output are found in the target output.By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.