mirror of
https://github.com/TaterTotterson/microWakeWord-Trainer-Nvidia-Docker.git
synced 2026-08-12 07:55:33 -06:00
285 B
285 B
- Prevented unreadable audio files from freezing sample preparation by adding a bounded FFmpeg watchdog, safe cleanup, and visible normalization progress.
- Kept primary and guided STT results visible after captures are automatically sorted into positive or negative training samples.