The "sherpa-onnx" topic on GitHub groups 8 open-source projects in the DeepSeek Harness (DSH) ecosystem, led by dsh-voice-mode with 19 GitHub stars. dsh-voice-mode — DSH full-duplex voice conversation mode: streaming zipformer2 recognition into an editable draft, optional wake word, Edge TTS sentence-by-sentence reading + real-time subtitles, barge-in on speech, no API Key needed. Every project here is indexed by DSH Universe with live GitHub data — stars, activity and install status — so you can compare and install directly.
DSH full-duplex voice conversation mode: streaming zipformer2 recognition into an editable draft, optional wake word, Edge TTS sentence-by-sentence reading + real-time subtitles, barge-in on speech, no API Key needed. Full-duplex voice mode for DeepSeek Harness, no API key.
DeepSeek Harness (dsh) voice input plugin: record with mic button/hotkey, real-time transcription backfilled into the input box. Three recognition engines: browser Web Speech, local SenseVoice/Paraformer offline backend (one-click deploy), OpenAI-compatible cloud ASR API.
DSH 0.2 桌面版语音插件(dsh-voice-mode 的非官方兼容分支):流式识别入草稿、按句朗读 + 实时字幕、开口即打断,另含 ASR 纠错词表与整段丢字修复。 · Unofficial dsh-voice-mode port for the DSH 0.2 desktop app: streaming ASR into the draft, per-sentence TTS with live captions, true barge-in, plus a local ASR correction dictionary.