#4435 TaskCfgVTT(is_cuda=True, uuid='92ecec3cbd', cache_folder='D:/shipinzhuanhuan/tmp/22048/92ecec3cbd', target_dir='E:/1688选

2409:8a44* Posted at: 2 hours ago 👁7

语音识别阶段出错 [faster-whisper(本地)] Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 347, in faster_whisper
File "faster_whisper\transcribe.py", line 689, in init
ValueError: Requested int8_float16 compute type, but the target device or backend do not support efficient int8_float16 computation.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 358, in faster_whisper
File "faster_whisper\transcribe.py", line 689, in init
ValueError: Requested float16 compute type, but the target device or backend do not support efficient float16 computation.

Traceback (most recent call last):
File "videotrans\task\job.py", line 105, in run
File "videotrans\task\trans_create.py", line 380, in recogn
File "videotrans\recognition\__init__.py", line 293, in run
File "videotrans\recognition\_base.py", line 143, in run
File "videotrans\recognition\_overall.py", line 33, in _exec
File "videotrans\recognition\_overall.py", line 105, in _faster
File "videotrans\configure\_base.py", line 289, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 347, in faster_whisper
File "faster_whisper\transcribe.py", line 689, in init
ValueError: Requested int8_float16 compute type, but the target device or backend do not support efficient int8_float16 computation.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 358, in faster_whisper
File "faster_whisper\transcribe.py", line 689, in init
ValueError: Requested float16 compute type, but the target device or backend do not support efficient float16 computation.
TaskCfgVTT(is_cuda=True, uuid='92ecec3cbd', cache_folder='D:/shipinzhuanhuan/tmp/22048/92ecec3cbd', target_dir='E:/1688选品视频/TK商品视频/新建文件夹/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红-mp4', source_language='简体中文', source_language_code='zh-cn', source_sub='E:/1688选品视频/TK商品视频/新建文件夹/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红-mp4/zh-cn.srt', source_wav='D:/shipinzhuanhuan/tmp/22048/92ecec3cbd/remove_noise.wav', source_wav_output='E:/1688选品视频/TK商品视频/新建文件夹/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红-mp4/zh-cn.m4a', target_language='马来西亚语', target_language_code='ms', target_sub='E:/1688选品视频/TK商品视频/新建文件夹/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红-mp4/ms.srt', target_wav='D:/shipinzhuanhuan/tmp/22048/92ecec3cbd/target.wav', target_wav_output='E:/1688选品视频/TK商品视频/新建文件夹/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红-mp4/ms.m4a', name='E:/1688选品视频/TK商品视频/Gegebear/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红.mp4', noextname='Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红', basename='Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红.mp4', ext='mp4', dirname='E:/1688选品视频/TK商品视频/Gegebear', shound_del_name=None, translate_type=0, tts_type=0, volume='+0%', pitch='+0Hz', voice_rate='+0%', voice_role='No', voice_autorate=True, video_autorate=False, remove_silent_mid=False, align_sub_audio=True, detect_language='zh-cn', recogn_type=0, model_name='large-v3-turbo', shibie_audio=None, remove_noise=True, enable_diariz=False, nums_diariz=0, rephrase=0, fix_punc=False, subtitle_language=None, app_mode='biaozhun', subtitles='', targetdir_mp4='E:/1688选品视频/TK商品视频/新建文件夹/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红-mp4/Gegebear净感水光多用棒滋润水光秋冬纯欲显嫩持妆韩系腮红膏口红.mp4', novoice_mp4='D:/shipinzhuanhuan/tmp/22048/92ecec3cbd/novoice.mp4', is_separate=False, embed_bgm=True, instrument=None, vocal='D:/shipinzhuanhuan/tmp/22048/92ecec3cbd/vocal.wav', back_audio='', clear_cache=False, background_music=None, subtitle_type=1, only_out_mp4=False, recogn2pass=True, output_srt=0, copysrt_rawvideo=False)


=
system:Windows-10-10.0.22621-SP0
version:v3.99
frozen:True
language:zh
root_dir:D:/shipinzhuanhuan

1 Replies

这是由当前显卡或CUDA环境不支持所请求的 int8_float16 / float16 计算类型导致的。faster-whisper 在启用CUDA加速时默认尝试更高效的数据类型,但你的设备或后端未完全兼容。

解决方案(按顺序尝试):

  1. 修改CUDA数据类型为float32
    打开菜单:工具/选项高级选项 → 找到 语音识别参数 下的 CUDA数据类型,将其由 int8default 改为 float32,然后重试。
  2. 确认CUDA与驱动版本
    本软件需要CUDA 12.8 和 cuDNN 9.11 以上。请检查 nvcc -Vnvidia-smi,确保显卡驱动已更新至最新,否则安装匹配版本。
  3. 如果仍然失败,暂时关闭CUDA加速
    在主界面取消勾选 CUDA加速,使用CPU进行语音识别(虽然速度慢,但可完成识别)。
    注意:large-v3-turbo 模型较大,CPU处理可能较慢,建议改用 mediumsmall 模型。
  4. (可选)尝试禁用降噪
    你当前开启了降噪功能,可能会占用额外资源。在“设置更多参数”中取消勾选 降噪 看是否改善。

请查阅相关文档:

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.

Related resource