Commit Graph

Select branches

Hide Pull Requests

feat/question-gen-v3

main

Mono Color

• cd0805aad6 chore: fix N806 lint - rename _RESAMPLE_VIDEO_INTERVAL to lowercase iomgaa 2026-07-14 06:11:01 -04:00
• 7c7ad67d81 fix(pipeline): move extra_gates after standard gate pass check iomgaa 2026-07-14 06:08:52 -04:00
• 36c712defa feat(pipeline): replace QuestionFamilySpec with TaskTypeStrategy iomgaa 2026-07-14 06:05:23 -04:00
• b9616de21e fix(tests): update test_generator_v2 to use new generator signatures iomgaa 2026-07-14 05:55:08 -04:00
• afa77173e3 refactor(question_gen): adapt generator/gates/store signatures for strategy iomgaa 2026-07-14 05:50:47 -04:00
• c49d0ff12f fix(sampler): validate level param rejects invalid values iomgaa 2026-07-14 05:43:28 -04:00
• e2325b6535 refactor(sampler): replace family_spec param with level+constraint iomgaa 2026-07-14 05:38:14 -04:00
• b6b6a48503 feat(question_gen): add TaskTypeStrategy Protocol and BaseTaskTypeStrategy iomgaa 2026-07-14 05:24:29 -04:00
• 832838350a docs: add TaskTypeStrategy design + Plan A framework plan iomgaa 2026-07-14 05:15:28 -04:00
• 9ee37a8534 fix(loader): use video_id from JSON data instead of filename iomgaa 2026-07-14 01:02:55 -04:00
• dec7346da3 feat(harness): add Action Recognition training experiment iomgaa 2026-07-14 00:58:54 -04:00
• 37d4519905 chore: lint and format per-category pool strategy implementation iomgaa 2026-07-12 22:58:10 -04:00
• 671db2f88c test(integration): add PerCategoryPoolStrategy end-to-end test iomgaa 2026-07-12 22:56:42 -04:00
• 5563d0d675 feat(cli): wire train mode with PoolStrategy selection and task_types in RunConfig iomgaa 2026-07-12 22:56:18 -04:00
• c66a00c924 feat(harness): refactor build_or_load_pools to accept PoolStrategy + per_category freeze format iomgaa 2026-07-12 22:53:19 -04:00
• e5b07ac974 feat(harness): add task_types, pool_split_mode, train_ratio, test_questions to RunConfig iomgaa 2026-07-12 22:46:28 -04:00
• 73ae1f7143 fix(harness): change _runs INSERT OR IGNORE to ON CONFLICT DO UPDATE for incremental infer iomgaa 2026-07-12 22:45:18 -04:00
• 8b48005a17 config: add pool_split_mode, train_ratio, test_questions to default.yaml iomgaa 2026-07-12 22:43:43 -04:00
• 21c6a53aed feat(harness): add PerCategoryPoolStrategy with correctness-stratified 2:1 split iomgaa 2026-07-12 22:41:48 -04:00
• ec4cbbdd44 style(core): move Path import to TYPE_CHECKING block in types.py iomgaa 2026-07-12 22:36:00 -04:00
• cd5c9c01fb feat(app): add PoolStrategy Protocol to application ports iomgaa 2026-07-12 22:34:47 -04:00
• 0d0f275134 feat(core): add PoolConfig dataclass for pool strategy configuration iomgaa 2026-07-12 22:33:49 -04:00
• 4fb7a61f8b fix(question_gen): resolve pipeline integration issues from final review iomgaa 2026-07-12 00:14:12 -04:00
• eecb86e27a feat(question_gen): add generate-v2 CLI subcommand and experiment script iomgaa 2026-07-12 00:04:42 -04:00
• f46e87258c refactor(question_gen): extract helpers to reduce pipeline_v2 CC below grade C iomgaa 2026-07-11 23:57:35 -04:00
• 206c553143 feat(question_gen): add v2 pipeline with retry loop and heavy check iomgaa 2026-07-11 23:54:21 -04:00
• 6d6eb8e3a3 refactor(question_gen): extract _validate_parsed_fields to reduce CC iomgaa 2026-07-11 23:43:35 -04:00
• cf51d2de9d feat(question_gen): add v2 generator with per-family prompt templates iomgaa 2026-07-11 23:41:36 -04:00
• 9053233f99 fix(question_gen): gates reject_reason returns raw reason; short-circuit skips LLM iomgaa 2026-07-11 23:33:57 -04:00
• 9f739e831d refactor(question_gen): reduce cyclomatic complexity in sampler_v2 iomgaa 2026-07-11 23:33:48 -04:00
• f74711cd11 feat(question_gen): add v2 material sampler with family constraints iomgaa 2026-07-11 23:30:56 -04:00
• 271d1682c9 feat(question_gen): add lightweight 4-gate quality check iomgaa 2026-07-11 23:30:13 -04:00
• 9627ac9cf9 fix(question_gen): raise ValueError on UPDATE of missing rows iomgaa 2026-07-11 23:23:48 -04:00
• 7abe92eb1c fix(question_gen): check_verbatim covers question_text + add missing blacklist patterns iomgaa 2026-07-11 23:23:37 -04:00
• 9525726133 fix(question_gen): rename finished_at to ended_at per schema spec iomgaa 2026-07-11 23:18:52 -04:00
• 74686dde68 refactor(question_gen): reduce check_forbidden_material complexity to A(3) iomgaa 2026-07-11 23:16:34 -04:00
• 6a5424a618 feat(question_gen): add SQLite run store for generation telemetry iomgaa 2026-07-11 23:15:53 -04:00
• c83d771923 feat(question_gen): add deterministic postprocess layer iomgaa 2026-07-11 23:14:43 -04:00
• 75e6d8c550 feat(question_gen): add 5 question family specs with sampling constraints iomgaa 2026-07-11 23:14:22 -04:00
• 811ffa648b feat(types): extend GeneratedQuestion with skill_target & difficulty_steps iomgaa 2026-07-11 23:05:38 -04:00
• 043d4aa46f docs(plans): add Spec-3 question-gen-v2 implementation plan iomgaa 2026-07-11 22:57:47 -04:00
• 8c9adfd3fa docs(wiki): record Spec-2 harness-eval finding (N/A — pure tooling addition) iomgaa 2026-07-11 12:13:22 -04:00
• a51b3c4238 fix(tools): validate concurrency params to prevent zero-semaphore deadlock iomgaa 2026-07-11 12:11:48 -04:00
• 0290005129 fix(scripts): align build_trees.sh with project shell conventions iomgaa 2026-07-11 12:07:14 -04:00
• 928ec8c37b feat(scripts): add batch tree build entry with smoke-test guide iomgaa 2026-07-11 12:04:34 -04:00
• c5bf7e1a61 fix(tools): widen error isolation to cover full single-video pipeline iomgaa 2026-07-11 12:02:21 -04:00
• 56fe138a46 feat(tools): batch tree build orchestration with shared API semaphore iomgaa 2026-07-11 11:56:36 -04:00
• 978ddef91b feat(tools): add build_trees skeleton with discovery helpers iomgaa 2026-07-11 11:41:22 -04:00
• e9073bfdc2 feat(tree): expose build_async and accept injected API semaphore iomgaa 2026-07-11 11:14:45 -04:00
• 25f5537974 docs(plans): add Spec-2 batch tree build implementation plan iomgaa 2026-07-11 11:11:02 -04:00
• 1d2385291a docs(wiki): record Spec-1 harness-eval finding (all metrics pass) iomgaa 2026-07-11 09:33:53 -04:00
• 40b156cf69 docs(wiki): record final Codex PASS for recipe redesign round iomgaa 2026-07-11 09:08:27 -04:00
• ebbd112d50 fix(tree): move Iterable import into type-checking block iomgaa 2026-07-11 09:05:04 -04:00
• a046862ca1 docs(wiki): record build-tree recipe redesign and diagnosis stage tags iomgaa 2026-07-11 09:04:57 -04:00
• f0284199b8 refactor(tree): narrow _clean_join parameter type to Iterable iomgaa 2026-07-11 09:04:25 -04:00
• 1c21e215e2 test(tree): cover L1 entity fields and block ordering in view_node iomgaa 2026-07-11 09:02:03 -04:00
• a2c6a97621 refactor(tree): extract _clean_join to keep entity-fields complexity at B iomgaa 2026-07-11 08:56:17 -04:00
• 4e0e05210d feat(search): append raw entity fields after view_node summary iomgaa 2026-07-11 08:53:56 -04:00
• 291a8108e1 test(agent): tighten type annotations in step-retry tests iomgaa 2026-07-11 08:48:04 -04:00
• badfcce4cb fix(agent): validate non-empty retry delays; pin exhaustion semantics iomgaa 2026-07-11 08:44:47 -04:00
• e3184c11f9 feat(agent): step-level retry for transient LLM errors (20s/40s backoff) iomgaa 2026-07-11 08:37:35 -04:00
• cf529f2c8f test(agent): add docstring to empty-content boundary test iomgaa 2026-07-11 08:33:36 -04:00
• 439dc29b3b fix(agent): reject argless action in normalization; add boundary test iomgaa 2026-07-11 08:28:59 -04:00
• 6034d4172d test(question_gen): update __all__ assertion to include synthesizer exports iomgaa 2026-07-11 08:21:06 -04:00
• a31b1fbf37 test(question_gen): drop tests for removed calibrate pairing validator iomgaa 2026-07-11 08:19:45 -04:00
• 8d84d5e236 fix(agent): normalize fenced and flat-args LLM outputs in parser iomgaa 2026-07-11 08:16:28 -04:00
• 39c6352781 docs(plans): add Spec-1 implementation plan with Codex review fixes iomgaa 2026-07-11 08:12:07 -04:00
• ef402c46a2 docs(designs): apply Codex review fixes to Spec-1/2/3 iomgaa 2026-07-11 07:54:48 -04:00
• a5666b4f16 docs(wiki): record tree-figure style unification and review fixes iomgaa 2026-07-11 07:52:22 -04:00
• a79c2ec753 docs(wiki): add three-round analysis findings and Spec-1/2/3 designs iomgaa 2026-07-11 07:45:51 -04:00
• 658e62054e feat(scripts): add generated-questions infer and calibrate entry iomgaa 2026-07-11 07:45:42 -04:00
• da70eb6e23 refactor(question_gen): slim calibrate to compare two existing runs iomgaa 2026-07-11 07:45:42 -04:00
• 307c64c388 fix(harness): infer mode honors CLI --questions over manifest iomgaa 2026-07-11 07:45:33 -04:00
• 735471df3c fix(question_gen): Information Synopsis 去重阈值放宽至 0.90 iomgaa 2026-07-10 08:55:47 -04:00
• 25c8d5ec42 fix(question_gen): 每次重试换视频避免同视频反复去重失败 iomgaa 2026-07-10 08:15:03 -04:00
• d0148e78ad docs(wiki): record main-figure final review verdict and caption notes iomgaa 2026-07-10 01:40:04 -04:00
• 6a6aae081b docs(wiki): record main-figure final layout and panel decisions iomgaa 2026-07-10 01:38:25 -04:00
• 69a4d17c0b config: LLM 总超时 120s→300s（5 分钟） iomgaa 2026-07-10 01:11:40 -04:00
• 5b51f4bd0c fix(synthesizer): generate_one 捕获所有异常避免 VLM 超时穿透崩溃 iomgaa 2026-07-10 01:08:03 -04:00
• 8d11513e54 fix: 禁止日志缓存，确保所有日志立刻输出 iomgaa 2026-07-09 23:15:02 -04:00
• c191be16e2 fix(question_gen): L3 frame_path 拼接视频目录前缀 iomgaa 2026-07-09 23:10:18 -04:00
• 09b2773375 config(scripts): generate_questions 默认并发 8→16 iomgaa 2026-07-09 23:07:57 -04:00
• c84de49dc4 feat(main): loguru 文件 sink 持久化运行日志 iomgaa 2026-07-09 22:56:34 -04:00
• 75016c4892 feat(scripts): generate_questions.sh 题目生成实验入口 iomgaa 2026-07-09 22:53:28 -04:00
• 4ebdf0b0ba fix(main): InfraSettings 默认值同步 retry_base=20 retry_max=120 iomgaa 2026-07-09 22:16:22 -04:00
• 1f83da9eb7 fix: embedding 改 cuda + OCR 超时 300s + 重试退避 20s 起 + jitter 防雷群 iomgaa 2026-07-09 22:14:57 -04:00
• 38a855b33e fix(scripts): 加 HF_HUB_OFFLINE 跳过在线检查 iomgaa 2026-07-09 13:56:25 -04:00
• d6ae0d85cd fix: Redis 缓存构造修正 + TTL=0 永不过期支持 iomgaa 2026-07-09 13:52:33 -04:00
• 953fb7a456 feat(scripts): infer_video_mme.sh 推理实验入口 iomgaa 2026-07-09 13:40:31 -04:00
• cabe5be038 fix(main): assemble_mode 从 'default' 修正为 'ids_expand' iomgaa 2026-07-09 13:21:05 -04:00
• f8f8b5b12d feat: main.py Composition Root（仅 infer 模式） iomgaa 2026-07-09 12:27:13 -04:00
• ba400417e2 config: concurrency=24, max_steps=40, breaker_threshold=48 iomgaa 2026-07-09 12:27:05 -04:00
• a7ca6d15ed feat(harness): InferenceDepsRouter per-video 路由器 iomgaa 2026-07-09 12:20:32 -04:00
• f21bf345a6 feat(runner): 注入 tool_dispatch_factory/prompt_builder_factory + fail-fast iomgaa 2026-07-09 12:20:25 -04:00
• 924160c779 feat(ports): 新增 ToolDispatchFactory/PromptBuilderFactory Protocol iomgaa 2026-07-09 12:20:16 -04:00
• e7be42570d feat(store): skills/v1 初始集 — TRM4 精简 + TRM5 card 字段注入 iomgaa 2026-07-09 12:12:47 -04:00
• bb6dad7005 refactor(store): prompts 版本化目录重组 + skills/v1 骨架 + system.md L1 去 subtitle iomgaa 2026-07-09 12:07:19 -04:00
• 16140cafcc fix(tree): from_dict L2 subtitle 旧格式兼容 fallback iomgaa 2026-07-09 12:04:36 -04:00
• d3be9b1322 refactor(tree): subtitle 迁入 L3Card/L2Card + 建树管线修正 iomgaa 2026-07-09 11:57:41 -04:00
• c72b55508a fix(plan): Task 0 Codex 审查修复 — 锚标保持/双写消除/幂等迁移/遗漏文件 iomgaa 2026-07-09 11:43:33 -04:00