#3891 TaskCfgVTT(is_cuda=True, uuid='11965ef7a8', cache_folder='E:/SP/tmp/9788/11965ef7a8', target_dir='E:/迅雷云盘/NTR/_video_out

167.179* Posted at: 1 day ago 👁22

语音识别阶段出错 [faster-whisper(本地)] 出错了,可能内存或显存不足 Model:large-v3-turbo GPU0
Traceback (most recent call last):
File "videotrans\configure\_base.py", line 285, in _new_process
File "concurrent\futures\_base.py", line 458, in result
File "concurrent\futures\_base.py", line 403, in __get_result
concurrent.futures.process.BrokenProcessPool: A process in the process pool was terminated abruptly while the future was running or pending.

Traceback (most recent call last):
File "videotrans\configure\_base.py", line 285, in _new_process
File "concurrent\futures\_base.py", line 458, in result
File "concurrent\futures\_base.py", line 403, in __get_result
concurrent.futures.process.BrokenProcessPool: A process in the process pool was terminated abruptly while the future was running or pending.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "videotrans\task\job.py", line 105, in run
File "videotrans\task\trans_create.py", line 360, in recogn
File "videotrans\recognition\__init__.py", line 272, 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 303, in _new_process
RuntimeError: 出错了,可能内存或显存不足 Model:large-v3-turbo GPU0
Traceback (most recent call last):
File "videotrans\configure\_base.py", line 285, in _new_process
File "concurrent\futures\_base.py", line 458, in result
File "concurrent\futures\_base.py", line 403, in __get_result
concurrent.futures.process.BrokenProcessPool: A process in the process pool was terminated abruptly while the future was running or pending.
TaskCfgVTT(is_cuda=True, uuid='11965ef7a8', cache_folder='E:/SP/tmp/9788/11965ef7a8', target_dir='E:/迅雷云盘/NTR/_video_out/NSFS-091-U-mp4', source_language='日语', source_language_code='ja', source_sub='E:/迅雷云盘/NTR/_video_out/NSFS-091-U-mp4/ja.srt', source_wav='E:/SP/tmp/9788/11965ef7a8/ja.wav', source_wav_output='E:/迅雷云盘/NTR/_video_out/NSFS-091-U-mp4/ja.m4a', target_language='简体中文', target_language_code='zh-cn', target_sub='E:/迅雷云盘/NTR/_video_out/NSFS-091-U-mp4/zh-cn.srt', target_wav='E:/SP/tmp/9788/11965ef7a8/target.wav', target_wav_output='E:/迅雷云盘/NTR/_video_out/NSFS-091-U-mp4/zh-cn.m4a', name='E:/迅雷云盘/NTR/NSFS-091-U.mp4', noextname='NSFS-091-U', basename='NSFS-091-U.mp4', ext='mp4', dirname='E:/迅雷云盘/NTR', shound_del_name=None, translate_type=5, tts_type=0, volume='+0%', pitch='+0Hz', voice_rate='+0%', voice_role='No', voice_autorate=False, video_autorate=False, remove_silent_mid=False, align_sub_audio=True, detect_language='ja', recogn_type=0, model_name='large-v3-turbo', shibie_audio=None, remove_noise=False, enable_diariz=False, nums_diariz=0, rephrase=0, fix_punc=False, subtitle_language=None, app_mode='tiqu', subtitles='', targetdir_mp4='E:/迅雷云盘/NTR/_video_out/NSFS-091-U-mp4/NSFS-091-U.mp4', novoice_mp4='E:/SP/tmp/9788/11965ef7a8/novoice.mp4', is_separate=False, instrument=None, vocal=None, back_audio='', clear_cache=True, background_music=None, subtitle_type=0, only_out_mp4=False, recogn2pass=True, output_srt=0, copysrt_rawvideo=True)


=
system:Windows-10-10.0.26200-SP0
version:v3.98
frozen:True
language:zh
root_dir:E:/SP

1 Replies

这是由于显存或内存不足导致 faster-whisper(本地)large-v3-turbo 模型加载失败。

解决方案:

  1. 更换更小的模型:在语音识别渠道中,将模型从 large-v3-turbo 更换为 mediumsmallbaselarge-v3-turbo 模型对显存/内存要求较高。
  2. 调整高级设置以降低资源占用:进入 菜单 -> 工具 -> 高级选项,在“语音识别参数”区域进行以下调整:

    • CUDA数据类型:将 float32 改为 float16int8
    • 识别准确度beam_size:将 5 改为 1
    • 识别准确度best_of:将 5 改为 1
    • 启用上下文感知:确保此项为 false(未勾选)。
  3. 检查并优化CUDA环境:确保已正确安装 CUDA 12.8 和 cuDNN 9.x,并在主界面勾选了“CUDA加速”。
  4. 关闭二次识别:在主界面右上角,取消勾选“二次识别”选项,尤其是在批量翻译时,可以避免同时运行多个识别任务导致显存需求翻倍。
  5. 确保软件路径合规:检查软件安装路径(E:/SP)和视频文件路径(E:/迅雷云盘/NTR)是否过长或包含中文字符。建议将软件和待处理视频都移动到更浅的英文路径下(如 D:\pyVideoTransD:\videos)。

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