#4763 TaskCfgSTT(uuid='850c81b7b9', name='N:/K/flv/media - Pleasure.mp4', dirname='N:/K/flv/', noextname='Pleasure, basename='

124.165* Posted at: 8 hours ago

语音识别阶段出错[faster-whisper(本地)] mkl_malloc: failed to allocate memory,Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 208, in faster_whisper
File "videotrans\process\stt_fun.py", line 173, in _create_model
File "faster_whisper\transcribe.py", line 689, in init
RuntimeError: mkl_malloc: failed to allocate memory

Traceback (most recent call last):

File "videotrans\task\job.py", line 54, in run

File "videotrans\task\job.py", line 119, in process_task

File "videotrans\task\speech2text.py", line 125, in recogn

File "videotrans\recognition\__init__.py", line 188, in run

File "videotrans\recognition\_base.py", line 93, in run

File "videotrans\recognition\_whisper.py", line 35, in _exec

File "videotrans\recognition\_whisper.py", line 108, in _faster

File "videotrans\configure\base.py", line 240, in _new_process

videotrans.configure.excepts.VideoTransError: mkl_malloc: failed to allocate memory,Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 208, in faster_whisper
File "videotrans\process\stt_fun.py", line 173, in _create_model
File "faster_whisper\transcribe.py", line 689, in init
RuntimeError: mkl_malloc: failed to allocate memory
TaskCfgSTT(uuid='850c81b7b9', name='N:/K/flv/media - Pleasure.mp4', dirname='N:/K/flv/', noextname='Pleasure, basename='Pleasure.mp4', ext='mp4', target_dir='O:/win-pyvideotrans-v4.00-525/output/recogn', cache_folder='O:/win-pyvideotrans-v4.00-525/tmp/896/850c81b7b9', is_cuda=False, 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='O:/win-pyvideotrans-v4.00-525/output/recogn/Pleasure.srt', target_wav=None, target_wav_output=None, detect_language='en', recogn_type=0, model_name='large-v3', shibie_audio='O:/win-pyvideotrans-v4.00-525/tmp/896/850c81b7b9/leasure -1779956370.0213454.wav', remove_noise=True, enable_diariz=False, nums_diariz=0, rephrase=0, fix_punc=False)


=
system:Windows-10-10.0.26100-SP0
version:v4.00
frozen:True
language:zh
root_dir:O:/win-pyvideotrans-v4.00-525

1 Replies

这是由于系统内存不足(或显存不足)无法加载 large-v3 模型导致的。您的 is_cuda=False 表明未使用显卡加速,此时 faster-whisper 依赖 CPU 和系统内存,而 large-v3 模型在 CPU 模式下需要大量 RAM(通常建议 16GB 以上)。

解决方法:

  1. 切换为更小模型:在语音识别渠道中将模型从 large-v3 改为 mediumsmall(显存/内存占用更小)。
  2. 调整高级参数:菜单栏 工具/选项高级选项语音识别参数,将:

    • CUDA数据类型 改为 int8
    • beam_size 改为 1
    • best_of 改为 1
    • 启用上下文感知 取消勾选
  3. 释放系统内存:关闭其他占用内存的程序后再运行。
  4. 若仍有不足:可考虑物理升级内存,或在支持 CUDA 的 NVIDIA 显卡上勾选主界面的 CUDA加速 来使用显存(需已安装 CUDA 12.8+ 和 cuDNN 9.x+)。

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