committed patch mode - #8
Merged
Merged
Conversation
Some SWE task families end their problem statement with an instruction to work on a new branch and commit everything when done. The agent obeys, leaving a clean working tree, so the existing `git diff` capture returns nothing and the rollout is recorded as a 0-byte patch. What non-empty patches do survive tend to be uncommitted build artifacts (build/CMakeFiles, *.o, *.a, .ninja_log) rather than the solution. Add a second capture mode, selected by --patch-mode (gym passes it through the PATCH_MODE env var). `worktree` is the default and is the historical behaviour moved verbatim out of cli.ts. `committed` diffs the pre-agent HEAD against the most-advanced commit the agent left, searching HEAD *and* every local branch so work still lands when the agent commits on a side branch and checks main back out. Uncommitted leftovers are excluded and logged. The result is still a plain baseline -> final-tree diff, so the eval side (git reset --hard <base_commit> + git apply) is unchanged in either mode. Also sets a repo-local git identity in committed mode when the image ships none, so the agent's first commit can't die with "Author identity unknown". Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
merge sdd/dev to compaction
|
Hey! Your PR title Please update it to start with one of:
Where See CONTRIBUTING.md for details. |
|
This PR doesn't fully meet our contributing guidelines and PR template. What needs to be fixed:
Please edit this PR description to address the above within 2 hours, or it will be automatically closed. If you believe this was flagged incorrectly, please let a maintainer know. |
…h-mode Signed-off-by: Sugam Devare <sdevare@nvidia.com>
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.
Issue for this PR
Closes #
Type of change
What does this PR do?
Please provide a description of the issue, the changes you made to fix it, and why they work. It is expected that you understand why your changes work and if you do not understand why at least say as much so a maintainer knows how much to value the PR.
If you paste a large clearly AI generated description here your PR may be IGNORED or CLOSED!
How did you verify your code works?
Screenshots / recordings
If this is a UI change, please include a screenshot or recording.
Checklist
If you do not follow this template your PR will be automatically rejected.