Zayden-Jung/lean-bench
2
1
Fork 0
Code Issues 2 Pull Requests Actions Packages Projects Releases Wiki Activity

Superseded by #9: Preserve P0 partial results and report the billing interruption #6

Closed
codex-bot wants to merge 0 commits from codex/pilot-execution into main
pull from: codex/pilot-execution
merge into: Zayden-Jung:main
Conversation 0 Commits - Files Changed -
+342 -41
codex-bot commented 2026-10-28 11:08:01 +00:00
Collaborator
Copy Link
Copy Source

Superseded by PR #9 at merge commit 3453d5c. The owner requested consolidating both open PRs into one. All commits from this branch are preserved in codex/simplex-registry, and combined CI #51 (API run ID 458) passed. This PR is closed without merging into main; the branch is retained. Review and manually merge PR #9 instead.

The owner requested the P0 execution fixes and report now, with paid resumption deferred after DashScope returned Arrearage. This follows the manually merged preparation PR #5 and contains only the six-file recovery/report change (16e49df).

The run issued 131 cumulative HTTP requests: 42 of 65 items have both detectors complete; 23 items still need 42 components. Preserve all 88 successful components. Recorded usage is 37,699 prompt tokens and 7,790 completion tokens. No GO/PARTIAL/NO-GO decision is valid; the scorer rejected the incomplete outputs.

Changes: stop immediately on provider billing/authentication errors; preserve successful components of partial items; replace rows atomically on resume; reconcile completion state with --audit-only. Add an English execution report and a text-free audit containing counts and hashes. Frozen labels, model, prompts, matching thresholds and concept tables remain unchanged.

Validation: 149 offline unit tests, repository-wide Ruff and git diff --check passed. Remote CI runs on the owner-confirmed ubuntu24.04-x64-lean4.35 runner. No local Lean or Docker build was run. Raw passages and model responses remain in the gitignored data directory.

Remaining work: restore provider billing, resume only missing components within the existing 400-request cumulative cap, then score all 65 items. Related issue #3 remains open. This PR is WIP because the full pilot is incomplete. Review and merge manually; no automatic merge is enabled.

Superseded by PR #9 at merge commit 3453d5c. The owner requested consolidating both open PRs into one. All commits from this branch are preserved in codex/simplex-registry, and combined CI #51 (API run ID 458) passed. This PR is closed without merging into main; the branch is retained. Review and manually merge PR #9 instead. --- The owner requested the P0 execution fixes and report now, with paid resumption deferred after DashScope returned Arrearage. This follows the manually merged preparation PR #5 and contains only the six-file recovery/report change (16e49df). The run issued 131 cumulative HTTP requests: 42 of 65 items have both detectors complete; 23 items still need 42 components. Preserve all 88 successful components. Recorded usage is 37,699 prompt tokens and 7,790 completion tokens. No GO/PARTIAL/NO-GO decision is valid; the scorer rejected the incomplete outputs. Changes: stop immediately on provider billing/authentication errors; preserve successful components of partial items; replace rows atomically on resume; reconcile completion state with --audit-only. Add an English execution report and a text-free audit containing counts and hashes. Frozen labels, model, prompts, matching thresholds and concept tables remain unchanged. Validation: 149 offline unit tests, repository-wide Ruff and git diff --check passed. Remote CI runs on the owner-confirmed ubuntu24.04-x64-lean4.35 runner. No local Lean or Docker build was run. Raw passages and model responses remain in the gitignored data directory. Remaining work: restore provider billing, resume only missing components within the existing 400-request cumulative cap, then score all 65 items. Related issue #3 remains open. This PR is WIP because the full pilot is incomplete. Review and merge manually; no automatic merge is enabled.
codex-bot added 1 commit 2026-10-28 11:08:01 +00:00
fix(pilot): preserve partial results and stop on provider billing errors
ci / python (pull_request) Successful in 17s
Details
ci / lean (pull_request) Successful in 1m21s
Details
ci / image (pull_request) Has been skipped
Details
16e49df764
codex-bot added 1 commit 2026-10-28 11:08:49 +00:00
docs(pilot): link the recovery audit to the follow-up PR
ci / python (pull_request) Successful in 13s
Details
ci / lean (pull_request) Successful in 27s
Details
ci / image (pull_request) Has been skipped
Details
1c265609d5
codex-bot changed title from WIP: Preserve P0 partial results and report the billing interruption to Superseded by #9: Preserve P0 partial results and report the billing interruption 2026-10-28 13:50:59 +00:00
codex-bot closed this pull request 2026-10-28 13:51:00 +00:00
Some checks are pending
ci / python (pull_request) Successful in 13s
Details
ci / lean (pull_request) Successful in 27s
Details
ci / image (pull_request) Has been skipped
Details

Pull request closed

Please reopen this pull request to perform a merge.
Sign in to join this conversation.
Reviewers
No Reviewers
Labels
No items
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
claude-bot codex-bot yilin-codex-bot Zayden-Jung
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Zayden-Jung/lean-bench#6
Delete Branch "codex/pilot-execution"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?