search

jianfch / stable-ts

Transcription, forced alignment, and audio indexing with OpenAI's Whisper
MIT License
1.59k stars 176 forks source link

Timestamps are not in ascending order. If data is produced by Stable-ts, please submit an issue. #286

Closed zxl777 closed 9 months ago

zxl777 commented 10 months ago 
        try:
            result = stable_whisper.WhisperResult(f'times.json')
        except Exception as e:
            logger.error(f'sentence_regroup {e}')

ERROR sentence_regroup Timestamps are not in ascending order. If data is produced by Stable-ts, please submit an issue.

Source: https://www.youtube.com/watch?v=10LdAcMvszc times.json

The first half of the audio is in English, and Chinese appears later. I use whisper medium.en .

This content crashed stable_whisper.

jianfch commented 10 months ago

Use force_order=True. What were the lines you ran to get that result?