extraction-workflows
Plan and troubleshoot VSE hard-subtitle extraction workflows, including source CLI runs, subtitle areas, modes, VideoSubFinder, outputs, caches, and progress behavior.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- extraction-workflows
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Plan and troubleshoot VSE hard-subtitle extraction workflows, including source CLI runs, subtitle areas, modes, VideoSubFinder, outputs, caches, and progress behavior.
- disable-model-invocation
- true
02 资源文件 5
- 参考文档references/api-reference.md2.2 KB
- 参考文档references/troubleshooting.md1.5 KB
- 参考文档references/videosubfinder.md1.6 KB
- 参考文档references/workflows.md3.2 KB
- 脚本scripts/vse_cli_plan.py2.8 KB
03 SKILL.md 正文
Extraction Workflows
Use this sub-skill when a user wants VSE to extract hard-coded subtitles from
video into .srt or .txt, run the source CLI, choose Fast/Auto/Accurate
modes, pass a subtitle area, understand VideoSubFinder, or debug empty/slow
extraction output.
Read first
- workflows: end-to-end extraction flow, inputs, outputs, cache behavior, and non-interactive planning.
- API reference:
SubtitleExtractormethods and runtime state that matter when integrating or debugging source runs. - VideoSubFinder: how Fast/Auto mode uses the bundled platform binary and what failures look like.
- troubleshooting: empty SRTs, wrong subtitle areas, hangs, path issues, and sampling/mode decisions.
- Run scripts/vse_cli_plan.py to build a safe plan for an extraction run without starting OCR.
Route decisions
- For OCR model/language/backend install details, route to ocr-backends.
- For typo maps, confidence threshold cleanup, duplicate-line handling, TXT output, or word segmentation, route to postprocessing-config.
- For GUI batch task setup or normalized preview rectangles, route to gui-batch-operations.
- For synchronizing an existing subtitle file to a different video, route to subtitle-sync.
Fast operating checklist
- Confirm the user has a VSE source checkout/release and a prepared Python environment. Use the root environment probe if uncertain.
- Confirm video path, desired output path or save directory, language code, mode, hardware acceleration preference, and subtitle area.
- Prefer Fast/Auto; choose Accurate only for missed subtitle intervals with an accepted runtime cost.
- If the subtitle area is unknown, ask for GUI selection or derive a pixel rectangle from a screenshot; do not run blindly unless the user accepts watermark/scene-text prompts.
- After extraction, route cleanup/formatting issues to the post-processing sub-skill.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.
passnet-solve
END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
passnet-pattern-fusion
Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).