mirror of
https://github.com/open-webui/open-webui.git
synced 2026-10-15 04:18:18 +00:00
86de14cff9
In a Realtime Voice call, any loud sound like wind on a phone or passing traffic stopped the assistant mid-sentence as if you had spoken, and the answer was lost. The assistant now keeps talking through noise and only stops once what you said has been transcribed into words, so even a single "Ja" still stops it. The tradeoff is that talking over the assistant stops it a little later than before, after you pause and your words are transcribed. Fixes #32248 |
||
|---|---|---|
| .. | ||
| data | ||
| open_webui | ||
| .dockerignore | ||
| .gitignore | ||
| dev.sh | ||
| requirements-slim.txt | ||
| requirements.txt | ||
| start.sh | ||
| start_windows.bat | ||