iomgaa 02b8145b7f fix: fail loud on diag model fingerprint drift vs .env SEARCH_LLM_MODEL ...

Codex CHANGES_REQUESTED 复审：
- Critical：diag_fingerprint 用 config.diag.model，但 Phase 1 诊断 LLM 从 .env
  SEARCH_LLM_MODEL 构建，两者不一致会让信号以错误模型指纹落库，破坏可复现/
  resume/隔离。build_diagnosis_deps 新增 expected_model 参数，Phase 1 执行前
  fail loud 校验 config.model == settings.search_llm_model（附两值）。
- Minor：config/video_split.yaml diag.model 注释由 JUDGE_LLM_MODEL 更正为
  SEARCH_LLM_MODEL，与实现对齐。
- 补两个单测：模型不一致 fail loud + 缺凭证 fail loud。

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-07-15 13:20:51 -04:00

..

__init__.py

feat(harness): __init__.py public API + lint fixes

2026-07-07 13:45:48 -04:00

baseline_diagnosis.py

fix: dedup three-bucket projection + honest C3 error-handling note

2026-07-15 12:29:58 -04:00

baseline_run_log.py

feat: add steps_json-backed RunLog wrapper

2026-07-15 11:51:55 -04:00

batching.py

feat(batching): unit 粒度切分——pair 整锁 + 单元级分桶 + 非 AR 独立 rng

2026-07-15 06:46:28 -04:00

build_split.py

feat: wire end-to-end results-driven split with defensive asserts

2026-07-15 12:43:25 -04:00

checkpoint.py

feat(harness): checkpoint 存 unit_id 序列，断点续跑孪生对不拆

2026-07-15 08:22:18 -04:00

config.py

feat(harness): add Action Recognition training experiment

2026-07-14 00:58:54 -04:00

deps_router.py

feat(harness): InferenceDepsRouter per-video 路由器

2026-07-09 12:20:32 -04:00

factory.py

feat(harness): add factory.py — InferenceDeps dataclass + build_inference_deps

2026-07-09 05:36:07 -04:00

gate_ladder.py

feat(harness): 迁移 gate 信息量阶梯到 unit 粒度 + schema_version

2026-07-15 07:56:15 -04:00

inference.py

refactor: fail-loud unit grading and stricter orphan pair drop

2026-07-15 07:05:37 -04:00

log.py

fix(harness): change _runs INSERT OR IGNORE to ON CONFLICT DO UPDATE for incremental infer

2026-07-12 22:45:18 -04:00

momentum.py

feat(harness): momentum.py — async 慢更新动量生成

2026-07-07 13:01:04 -04:00

observation.py

test(harness): 补 _ladder_units 直测 + 澄清 gate 观测表 unit_id 口径注释

2026-07-15 07:40:36 -04:00

pools.py

feat: add video-split config knobs and reproducible script

2026-07-15 12:54:25 -04:00

question_units.py

feat(harness): correctness 三口径 + gate 块按 unit 跑

2026-07-15 07:31:03 -04:00

runner.py

feat(harness): checkpoint 存 unit_id 序列，断点续跑孪生对不拆

2026-07-15 08:22:18 -04:00

split_manifest.py

fix: make pools.json freeze atomic + add split manifest

2026-07-15 12:28:48 -04:00

split_selection.py

feat: add video-split config knobs and reproducible script

2026-07-15 12:54:25 -04:00

steps_json_traces.py

feat: add steps_json to trace-row adapter

2026-07-15 11:47:00 -04:00

store.py

feat(harness): store.py — Store 版本操作 + Seed 管理

2026-07-07 12:36:36 -04:00

validate.py

feat(harness): correctness 三口径 + gate 块按 unit 跑

2026-07-15 07:31:03 -04:00

video_split_cli.py

fix: fail loud on diag model fingerprint drift vs .env SEARCH_LLM_MODEL

2026-07-15 13:20:51 -04:00

workspace.py

feat(harness): workspace.py — Workspace lifecycle + VersionedSkillStore/PromptStore

2026-07-07 12:41:59 -04:00