Skip to content

Remove duplicate proof-completion tasks - #90

Open
Qian-Cheng-nju wants to merge 4 commits into
mainfrom
cleanup-proof-completion-duplicates
Open

Remove duplicate proof-completion tasks#90
Qian-Cheng-nju wants to merge 4 commits into
mainfrom
cleanup-proof-completion-duplicates

Conversation

@Qian-Cheng-nju

@Qian-Cheng-nju Qian-Cheng-nju commented Jul 30, 2026

Copy link
Copy Markdown
Member

Summary

  • remove 23 byte-identical proof-completion target copies while retaining canonical tasks
  • add one post-generation duplicate gate that auto-removes the approved copied-module families listed in JSON and rejects unknown duplicate groups before deleting anything
  • update dataset counts to 68 examples and 683 tasks

@Qian-Cheng-nju
Qian-Cheng-nju marked this pull request as ready for review July 30, 2026 14:30
@Qian-Cheng-nju

Copy link
Copy Markdown
Member Author

@sukanya1426 Could you review that? Thanks a lot!

@sukanya1426

Copy link
Copy Markdown
Contributor

@Qian-Cheng-nju Sure..

Comment thread src/dataset/proof_completion/generate.py
Comment thread src/dataset/proof_completion/duplicate_task_families.json
Comment thread src/dataset/proof_completion/generate.py Outdated
@Qian-Cheng-nju

Copy link
Copy Markdown
Member Author

Just a heads-up since this overlaps with #92

@sukanya1426 Thanks for pointing this out. Since these duplicate groups are introduced by #92, I think the allowlist should be updated in that PR.

I have fixed other comments, thanks a lot!

@Qian-Cheng-nju

Copy link
Copy Markdown
Member Author

@sukanya1426 Once you think it looks good, please merge.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants