New
modelship-ai/modelship — Preserve fractional audio duration in transcription responses
A commit notice about a fix in modelship-ai/modelship where the audio duration in whispercpp transcription responses was truncated to whole seconds, now preserving the fractional part.