{"commit":{"body":[{"label":"Why","text":"학습주기가남아도반복되던score/NMS비용을줄이고현재시나리오On처리량과송신손실경계를구분한다."},{"label":"What","text":"메모리주기hint/초당설정갱신/authoritative예약보호와표적테스트,trainingworker계측및Pose복합기준선문서를추가했다."},{"label":"Impact","text":"학습조건/주기/용량,provider직접송신,BOff,batch8instance8,Event10초,Stream300유지.모델/SDK/자원튜닝없음."},{"label":"Verify","text":"Python3.12 training9PASS.Pose1-16완료,복합1-9완료.복합10확장중Pose650응답뒤stall로10-16미측정.46child exit0/unforced/reap,36stopped/설정복귀/모델4READY/WSS.성능회귀나물리한계미확정."},{"label":"Rollback","text":"이전image와전체privateCompose보존.명시승인뒤Stop/reap과image설정정합으로복귀하며학습/Event원본이나원장은삭제하지않는다.자동rollback/재Start없음."}],"title":"fix(training): 주기 대기 중 후보 검사를 줄이고 채널 기준선 기록"},"created_at":"2026-09-27T03:42:54+00:00","expires_at":"2026-09-28T03:42:54+00:00","job_id":"20260927-contextra-pose-mixed-baseline-r01","paths":[{"change":"modified","content_sha256":"28de755e0b1f05aaaa371fcde9bce26ae0b06b9aef78f04f73c44a5d4f2230c3","path":"app/training_capture.py"},{"change":"modified","content_sha256":"1e9dd35fe52514f94429919973d88db74c29bc8c334119249e97ebef20e64a76","path":"app/training_store.py"},{"change":"modified","content_sha256":"05af95f1aacca1fd3b9efa8bfef3da468bb2ab0fc59d7e5c85970c00a508da32","path":"docs/profiles/ltb1021/topics/runtime-lifecycle/README.ko.md"},{"change":"added","content_sha256":"5cc0abdd14f03ee237cf40a59153eecccb57fb771719cc2451ce3f9b68836623","path":"docs/profiles/ltb1021/topics/runtime-lifecycle/analysis/20260927T123649+0900-analysis-pose-mixed-baseline.md"},{"change":"added","content_sha256":"31c827d9ed77f8b92bcd4b1b90dda49ca9b528e3d9f6800304cf245baf420973","path":"docs/profiles/ltb1021/topics/runtime-lifecycle/validation/20260927T123649+0900-validation-pose-mixed-baseline.md"},{"change":"modified","content_sha256":"7a1b04747d1a8eece3fafbda5e3a53cbb1be67922eb4ba6a7d255d311e5eb02a","path":"docs/scenarios/PROVIDER_PROCESS_RUNTIME.ko.md"},{"change":"modified","content_sha256":"80efaf0813f15716e0aa27a20080265402170dd088ec2466d1ebc45436a7cb6b","path":"docs/scenarios/TRAINING_SNAPSHOTS.ko.md"},{"change":"modified","content_sha256":"dba55d633f8350b35502fe9f73603a7b9b79b34efa3aa82432931fb1242cda86","path":"tests/test_training.py"},{"change":"modified","content_sha256":"717a3c49bfa8bea618beeca4e4b5679ebb919366ab58e61d7fe542daa9043d36","path":"tools/provider_scale_probe.py"}],"plan_hash":"0482ee49068d81aa357a804c52fa1300b00a5bf1855689428f20c9b83130bf02","profile":"ltb1021","push":{"branch":"task/runtime-lifecycle-refactor","mode":"upstream","remote":"origin","remote_identity":"sha256:3a9ce41319a2ebea","upstream_head":"66b3510271ddf27d440f5b5566d6d05b85bedffb","upstream_ref":"origin/task/runtime-lifecycle-refactor"},"repository":{"baseline_head":"66b3510271ddf27d440f5b5566d6d05b85bedffb","branch":"task/runtime-lifecycle-refactor","worktree":"/workspace/contextra-analytics-runtime-lifecycle-refactor"},"schema_version":"ltb1021.codex_hub.git_closeout_manifest.v1","session_alias":"work-session","verify":["Python3.12 training9PASS; exact source/image hashes; git diff check and documentation links","Pose1-16 and mixed1-9 measured; mixed10 stall boundary preserved; 46 reaped; final settings/models/WSS"]}