• 2025-01-08通过whisper制作批量语音数据集
    importsubprocessimportosimportwhisperimportzhconvimportjsondeftime2ffmpeg_time(ti):h=int(ti//3600)ti%=3600m=int(ti//60)ti%=60s=int(ti)ms=f'{ti:.2f}'[-2:]returnf'{h:02d}:{m:02d}:{s:02d}.{ms}�