m-bain / whisperX

WhisperX: Automatic Speech Recognition with Word-level Timestamps (& Diarization)
BSD 2-Clause "Simplified" License
12.49k stars 1.31k forks source link

No punctuation for Chinese #645

Open wntg opened 10 months ago

wntg commented 10 months ago

When I recognized one minute of Chinese audio, I found that there was no punctuation

MinJ-lucky commented 8 months ago

same issue. any solution now?

bitiliu commented 7 months ago

yes, same issue, any solution now?

LiMengSir commented 4 months ago

me too,Is there a solution?

leexinxu commented 3 months ago

asr_options = { "initial_prompt": "以下是中文普通话句子。", } whisper_model = whisperx.load_model(model_name, download_root=download_root, device=device, compute_type='float32', asr_options=asr_options)