Transcribe local audio or video with Volcengine Doubao file ASR, including BigASR 1.0 Turbo direct upload and asynchronous 1.0 standard, 1.0 idle, or 2.0 standard jobs through TOS. Use when the user asks for 火山引擎、豆包语音、录音文件识别、音视频转文字、录音稿或字幕。
作为 `lecture-to-md` 的可选云端 ASR 子 skill；默认走本地 sherpa-onnx X-ASR，仅在用户明确要求云端 API 或需更高准确率时启用。
中文简介正在准备中。当前显示原文说明。
查看安装方法安装前可以查看交给 AI 代理的指令内容。
lecture-to-md 子 skill）Use the deterministic scripts in scripts/; do not reproduce API requests manually.
本 skill 是
lecture-to-md的云端 ASR 子 skill。日常请走local-asr/（sherpa-onnx X-ASR，免费、跨平台、隐私友好）；只有当用户明确要求云端准确率，或本地模型不足以应对时，才启用本 skill。
turbo-v1, run scripts/turbo_transcribe.py. It accepts a local audio/video file, extracts a 16 kHz mono MP3, sends Base64 in one synchronous request, and writes JSON/TXT/SRT/Markdown.standard-v1, idle-v1, or standard-v2, first run scripts/upload_submit.py, then run scripts/poll_transcript.py with the resulting job.json.standard-v2 when the user has opened 豆包录音文件识别模型 2.0.--audio-url and skip TOS.Read references/api-modes.md before changing endpoints, resource IDs, limits, or request fields.
Read credentials from .secret in this skill directory. Never print or copy credential values into results. .secret must remain gitignored and mode 0600.
The ASR v3 calls use only VOLCENGINE_ASR_APP_ID and VOLCENGINE_ASR_ACCESS_TOKEN. VOLCENGINE_ASR_SECRET_KEY is retained locally but not sent. TOS uses a separate account-level AK/SK, endpoint, region, and bucket.
When .venv is absent, prepare the optional TOS dependency once:
python3 -m venv .venv
.venv/bin/python -m pip install -r requirements.txt
upload_submit.py automatically re-executes with this environment for local TOS uploads.
Turbo 1.0:
python3 scripts/turbo_transcribe.py "/absolute/path/video.mp4"
Submit a local 2.0 standard job through TOS:
python3 scripts/upload_submit.py \
--mode standard-v2 \
--input "/absolute/path/video.mp4"
Poll the generated job:
python3 scripts/poll_transcript.py "/absolute/path/job.json"
For a pre-existing direct URL:
python3 scripts/upload_submit.py \
--mode standard-v2 \
--audio-url "https://example.com/audio.mp3"
seedasr Turbo or idle resource IDs.job.json when polling times out. Delete the TOS object only after a terminal success/failure unless the user passes --keep-upload.transcript.txt, transcript.srt, and result.json after success.还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
把课堂视频（本地或 B 站/YouTube）、文字稿、课件三者（任意组合）整理成一份详细的中文 Markdown 课堂笔记，输出按课程标题命名的 `{title_name}.md`（首行为 `# 文档标题`）+ 相对路径图片。Markdown 工作流，与上游 `lecture-to-notes` 的 LaTeX/PDF 输出并行存在；上游 skill 完全不动。触发词：markdown 笔记、md 笔记、视频转 markdown、Markdown 讲义、不要 LaTeX、不要 PDF、纯文本笔记。
中文简介正在准备中。当前显示原文说明。
Use when users provide YouTube, Bilibili, or X/Twitter lecture URLs and want reader-first Chinese LaTeX/PDF notes with source-faithful claims, fluent authored prose, and verified teaching figures, especially requests phrased as lecture notes, 视频转PDF, 课程笔记, 讲义, YouTube笔记, B站笔记, X/Twitter lecture notes, or BV号; do not produce transcript dumps, quota-padded prose, or screenshot galleries.
中文简介正在准备中。当前显示原文说明。
把本地长视频/音频转写成文字稿 + 可选字幕，纯本地（不上传云端），用 sherpa-onnx X-ASR Zipformer transducer 模型（int8 量化、中英双语、自动标点）。已在 macOS Apple Silicon（int8 + AMX，~100× 实时）、Linux ARM64（CPU，~32× 实时）与 Windows（PowerShell 5.1，CPU）端到端验证。使用 `$lecture-to-md` 默认 ASR 后端；默认不要换成 Qwen。
中文简介正在准备中。当前显示原文说明。
Generate a self-contained, beautifully styled HTML analysis of an academic paper. Use when the user provides an arXiv link, paper URL, PDF, or asks to analyze a research paper. Produces a standalone HTML file with structured sections (Problem, Translation/Analogy, Architecture, Key Results, Verdict), embedded figures, and responsive design. Trigger words include 读论文, 分析论文, paper analysis, paper review.
中文简介正在准备中。当前显示原文说明。