#4390 TaskCfgSTT(is_cuda=True, uuid='707d968762', cache_folder='E:/PyVideoTrans v3.99(Subtitle generation and translation soft

14.153* Posted at: 8 hours ago 👁13

语音识别阶段出错 [faster-whisper(本地)] Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 471, in faster_whisper
File "faster_whisper\transcribe.py", line 1851, in restore_speech_timestamps
File "faster_whisper\transcribe.py", line 1279, in generate_segments
File "faster_whisper\transcribe.py", line 1590, in add_word_timestamps
File "faster_whisper\transcribe.py", line 1709, in find_alignment
RuntimeError: cuBLAS failed with status CUBLAS_STATUS_NOT_SUPPORTED

Traceback (most recent call last):
File "videotrans\task\job.py", line 105, in run
File "videotrans\task\_speech2text.py", line 159, 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 471, in faster_whisper
File "faster_whisper\transcribe.py", line 1851, in restore_speech_timestamps
File "faster_whisper\transcribe.py", line 1279, in generate_segments
File "faster_whisper\transcribe.py", line 1590, in add_word_timestamps
File "faster_whisper\transcribe.py", line 1709, in find_alignment
RuntimeError: cuBLAS failed with status CUBLAS_STATUS_NOT_SUPPORTED
TaskCfgSTT(is_cuda=True, uuid='707d968762', cache_folder='E:/PyVideoTrans v3.99(Subtitle generation and translation software)/tmp/3036/707d968762', target_dir='E:/PyVideoTrans v3.99(Subtitle generation and translation software)/output/recogn', source_language=None, source_language_code=None, source_sub=None, source_wav=None, source_wav_output=None, target_language=None, target_language_code=None, target_sub='E:/PyVideoTrans v3.99(Subtitle generation and translation software)/output/recogn/Evolution of Sydney Fixed-Camera Timelapse - 2,000 Years in 10 Minutes.srt', target_wav=None, target_wav_output=None, name='J:/YouTube/DIY/Evolution of Sydney Fixed-Camera Timelapse - 2,000 Years in 10 Minutes.mp3', noextname='Evolution of Sydney Fixed-Camera Timelapse - 2,000 Years in 10 Minutes', basename='Evolution of Sydney Fixed-Camera Timelapse - 2,000 Years in 10 Minutes.mp3', ext='mp3', dirname='J:/YouTube/DIY', shound_del_name=None, detect_language='en', recogn_type=0, model_name='large-v3', shibie_audio='E:/PyVideoTrans v3.99(Subtitle generation and translation software)/tmp/3036/707d968762/Evolution of Sydney _ Fixed-Camera Timelapse - 2,000 Years in 10 Minutes-1777780214.7290614.wav', remove_noise=False, enable_diariz=False, nums_diariz=0, rephrase=0, fix_punc=False)


=
system:Windows-10-10.0.26200-SP0
version:v3.99
frozen:True
language:zh
root_dir:E:/PyVideoTrans v3.99(Subtitle generation and translation software)

1 Replies

这是由 NVIDIA 50 系列显卡(Blackwell 架构)与 CUDA 库不兼容导致的 cuBLAS 调用失败。
解决方案:

  1. 确认已安装 CUDA 12.8 和 cuDNN 9.11(与软件内置版本匹配)。
  2. 在软件菜单中进入 工具/选项 → 高级选项,将 CUDA数据类型 改为 float16int8
  3. 仍然不行时,在系统环境变量中添加以下项并重启软件:

    • 变量名:CUBLAS_WORKSPACE_CONFIG,值::4096:8
    • 变量名:NVIDIA_TF32_OVERRIDE,值:0
  4. 做为临时规避,可暂时取消勾选主界面的“CUDA加速”,改用 CPU 推理(速度会慢)。

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