WIP: Add Capture Dataset #974
Open
+191
−3
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
|
Hi, @bodsul , Maybe you can try to re-implement and use some API models (like gpt-4o / gpt-4.1-mini). If you don't have the credit for GPT API, you can just implement in this PR and I'll help check the results. |
|
Added my comments to the issue in the original repo, lmk if any other clarification is needed. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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Outdated suggestions cannot be applied.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.
Add evaluation for newly released CAPTURE dataset. Arxiv preprint: https://arxiv.org/abs/2504.15485, repo: https://github.com/atinpothiraj/CAPTURe.
Opened issue regarding prompt choice and using Llama-3.1-8B-Instruct to extract answers here. Will update PR from WIP once this issues are clarified.