shuke987 opened a new pull request, #69988:
URL: https://github.com/apache/doris/pull/69988
### What problem does this PR solve?
Related PR: #69953
When GPT-6.1 Sol encounters five capacity failures, review recovery
currently switches to GPT-6 Astra, which has continued to encounter capacity
failures in observed runs. Change the capacity fallback to `gpt-6-sol` and
update the existing test fixtures and assertions to exercise that target.
The production change is one workflow argument. The primary model remains
`gpt-6.1-sol` with `xhigh`; the five-failure threshold, seven total capacity
attempts, 300-second backoff, same-session recovery, shared budget, ledger and
round preservation, head-change checks, and submission verification are
unchanged.
### Release note
None
### Check List (For Author)
- Test:
- [x] Unit Test: `PYTHONDONTWRITEBYTECODE=1 python3 -m unittest discover
-s .github/scripts -p 'test_*.py' -v` — 323 tests, 301 passed and 22 skipped on
macOS; no failures.
- [x] Manual test: `bash -n` on the extracted review step, `git diff
--check`, and exact comparison confirming the single production argument change
and unchanged recovery helper.
- Existing tests cover switching after five failures, same-session and
budget preservation, child-agent recovery instructions, seven-attempt
exhaustion, non-capacity errors, target changes, and actual workflow shell
argument forwarding.
- Live CI-account/model capacity and end-to-end cross-model recovery have
not been revalidated by this change.
- Behavior changed:
- [x] Yes: capacity recovery uses GPT-6 Sol instead of GPT-6 Astra.
- Does this need documentation?
- [x] No.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
