Repository Issue Activity (beta)

#1468 FFmpeg and PyTorch problem after second run

#297 Silence with background noise causing transcript to misalign

#1467 Feature Request: Add NVIDIA NeMo Sortformer as an Optional Speaker Diarization Backend

#1382 error message whisperx python

#1374 When whisperx can suport torch==2.9.1

#1449 Fix for #1220 (PR #1367) was reintroduced by #1386 — wildcard alignment code is back in v3.8.3+

#298 A "phantom language" can ruin the whole transcription

#1440 Potential optimisation: temp-file mode for load_audio() to reduce memory on long audio

#1404 Requesting for whisperx for python 3.14

#1247 Word-level timestamps from WhisperX are inaccurate compared to Montreal Forced Aligner (MFA)

#1461 uv pip install whisperx error

#1459 fix(asr): post-filter high compression_ratio segments like WhisperX

#293 Accuracy of Transcription: Filler words

#1456 `SubtitlesWriter` long-term vision?

#1457 `hf_token` can only be passed as a CLI argument

#9 Enhancement: Possible to determine different speakers?

#1310 Why WhisperX's sentence segmentation not working for Chinese?

#1450 Errors downloading punkt_tab data are silenced

#1451 Aligner gives inaccurate timestamps for a 25s audio segment (chunk from a larger clip)

#1448 Камаз

#998 New warnings in v3.3.1 acceptable?

#1446 Alignment fails when the same words are used in succession

#1195 `assign_word_speakers` expects a pd.Dataframe but pyannote pipeline returns Annotation

#1244 Changes regarding numpy?

#1416 Why is using python calls much slower than using cli