You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A first implementation is now open in #1163: it adds an optional SenseVoice/FunASR ASR provider for both CLI and UI flows, keeps the heavy runtime behind the optional sensevoice extra, and hides the provider when funasr/torch/torchaudio are not installed.
Current head aeaa7de5e6cafab6655b0210920b64c6027a6f7f has local validation for the focused SenseVoice/provider tests, Ruff, Pyright on the touched ASR/validator files, Python compilation, git diff --check, and package build. The remaining GitHub-side gate is that fork workflow runs need maintainer approval before CI can execute.
Activity
philpw99 commentedon Feb 11, 2026
SenseVoice 主要应用是实时语音转录，如果要用到这里的话，需要有一个能生成准确的时间轴或者字幕的程序才行。
Faster Whisper 和 Whisper-CPP 都是独立的程序，而它们都可以直接从音频生成字幕文件。
LauraGPT commentedon Jul 19, 2026
A first implementation is now open in #1163: it adds an optional SenseVoice/FunASR ASR provider for both CLI and UI flows, keeps the heavy runtime behind the optional
sensevoiceextra, and hides the provider whenfunasr/torch/torchaudioare not installed.Current head
aeaa7de5e6cafab6655b0210920b64c6027a6f7fhas local validation for the focused SenseVoice/provider tests, Ruff, Pyright on the touched ASR/validator files, Python compilation,git diff --check, and package build. The remaining GitHub-side gate is that fork workflow runs need maintainer approval before CI can execute.