A more general solution to model answer extraction instead of output_regex
#358
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
sadra-barikbin
commented
Oct 11, 2024
| @@ -504,7 +507,19 @@ def get_metric_method_from_category(self, metric_category): | |||
| if not self.has_metric_category[metric_category]: | |||
| raise ValueError(f"Requested a metric category {metric_category} absent from the task list.") | |||
|
|
|||
| return LightevalTask._get_metric_method_from_category(metric_category) | |||
| metric_method = LightevalTask._get_metric_method_from_category(metric_category) | |||
| # Bad hack. I had no other way. | |||
There was a problem hiding this comment.
I suggest considering task as the argument of apply_*_metrics functions as a workaround.
|
Hi Sadra, |
NathanHB
reviewed
Oct 14, 2024
Co-authored-by: Nathan Habib <30601243+NathanHB@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Outdated suggestions cannot be applied.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.
Hi there!
Here is the PR for model answer extraction. Currently, we have
output_regexand this PR attempts to replace it with a more general solution.By the way, our current
output_regexseems to be broken as it's not fed intoapply_generative_metricinPipeline._compute_metricsFixes #360