Repository navigation
Conversation
…h GLM-5.3 Adds a complete worked example for optimizing a skill-package capability against SWE-bench Verified using the Harbor adapter and GLM-5.3: - capevolve.yaml: cap-evolve spec (hill-climb, skill-package/system-prompt, Harbor dataset adapter, paired gate) - run.sh: reproducible launcher for Harbor on OpenShift via an AI gateway, with env wiring for agent/optimizer and headless claude onboarding - run_full/: full 500-task run artifacts — baseline/final JSON scores, events/graph/history/rejected JSONL, and optimizer JOURNAL/INSIGHTS The run reached 0.976 val and 0.959 test reward after 10 candidates.
yyualvin
marked this pull request as draft
October 6, 2026 14:32
This was referenced Oct 11, 2026
Collaborator
|
@yyualvin Can you pls push it to this branch: benchmark_history ? |
Author
|
Hi @OsherElhadad sorry this is a residual PR, the actual one is merged #699 |
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds a complete worked example for optimizing a skill-package capability
against SWE-bench Verified using the Harbor adapter and GLM-5.3:
Harbor dataset adapter, paired gate)
with env wiring for agent/optimizer and headless claude onboarding
events/graph/history/rejected JSONL, and optimizer JOURNAL/INSIGHTS
The run reached 0.976 val and 0.959 test reward after 10 candidates.