#1621 语音识别阶段出错:[faster-whisper(本地)] 运行时错误:Traceback (most recent call last): File "videotrans\process\_overall.py", line 93,

103.97* Posted at: 1 day ago 👁19

语音识别阶段出错:[faster-whisper(本地)] 运行时错误:Traceback (most recent call last):
File "videotrans\process\_overall.py", line 93, in run
File "faster_whisper\transcribe.py", line 689, in init
ValueError: Requested bfloat16 compute type, but the target device or backend do not support efficient bfloat16 computation.
:
Traceback (most recent call last):
File "videotrans\task\job.py", line 113, in run
File "videotrans\task\trans_create.py", line 354, in recogn
File "videotrans\recognition\__init__.py", line 236, in run
File "videotrans\recognition\_base.py", line 78, in run
File "videotrans\recognition\_overall.py", line 193, in _exec
RuntimeError: Traceback (most recent call last):
File "videotrans\process\_overall.py", line 93, in run
File "faster_whisper\transcribe.py", line 689, in init
ValueError: Requested bfloat16 compute type, but the target device or backend do not support efficient bfloat16 computation.

TaskCfg(cache_folder='D:/videotrans/tmp22652/60ef6492d4', target_dir="E:/Iceberg President's Substitute Soldier King/_video_out/1", remove_noise=False, is_separate=False, detect_language='en', subtitle_language=None, source_language='英语', target_language='西班牙语', source_language_code='en', target_language_code='es', source_sub="E:/Iceberg President's Substitute Soldier King/_video_out/1/en.srt", target_sub="E:/Iceberg President's Substitute Soldier King/_video_out/1/es.srt", source_wav='D:/videotrans/tmp22652/60ef6492d4/en.wav', source_wav_output="E:/Iceberg President's Substitute Soldier King/_video_out/1/en.m4a", target_wav='D:/videotrans/tmp22652/60ef6492d4/target.wav', target_wav_output="E:/Iceberg President's Substitute Soldier King/_video_out/1/es.m4a", subtitles='', novoice_mp4='D:/videotrans/tmp22652/60ef6492d4/novoice.mp4', noextname='1', shibie_audio=None, targetdir_mp4="E:/Iceberg President's Substitute Soldier King/_video_out/1/1.mp4", instrument=None, vocal=None, back_audio='', background_music=None, app_mode='biaozhun', subtitle_type=2, volume='+0%', pitch='+0Hz', voice_rate='+0%', voice_role='Elena(Female/AR)', copysrt_rawvideo=False, clear_cache=True, translate_type=5, tts_type=0, recogn_type=0, model_name='large-v3-turbo', split_type=0, voice_autorate=True, video_autorate=True, cuda=True, name="E:/Iceberg President's Substitute Soldier King/1.mp4", basename='1.mp4', ext='mp4', dirname="E:/Iceberg President's Substitute Soldier King", shound_del_name=None, uuid='60ef6492d4', remove_silent_mid=False, align_sub_audio=True, enable_diariz=True, nums_diariz=0, auto_fix=True)


=
system:Windows-10-10.0.19045-SP0
version:v3.86
frozen:True
language:zh
root_dir:D:/videotrans

1 Replies

这个错误是由于 Faster-Whisper 请求使用 bfloat16 计算类型,但您的设备或后端不支持高效的 bfloat16 计算。请按以下步骤解决:

  1. 进入高级设置:点击菜单栏的 工具/选项 -> 高级选项
  2. 修改计算类型:在 "Faster/OpenAI/Whisper 语音识别微调" 区域,找到 "CUDA数据类型" 选项
  3. 选择兼容类型:将计算类型从 bfloat16 改为 float16int8
  4. 降低识别精度:同时将 "识别准确度 (beam_size)" 和 "识别准确度 (best_of)" 都改为 1 以进一步减少显存占用
  5. 应用设置:点击确定保存设置,然后重新运行语音识别

如果问题依旧,建议:

  • 尝试使用更小的模型(如 medium 而非 large-v3-turbo
  • 在高级选项中取消勾选 "启用上下文感知"
  • 确保您的 NVIDIA 显卡驱动是最新版本

这些设置调整将解决 bfloat16 兼容性问题并减少显存需求。

Post Your Reply
Open source and free maintenance is not easy. If this project is helpful to you, please consider making a small donation to help the project continue to maintain and update.

Donate: https://ko-fi.com/jianchang512

Trending Questions