#5177 TaskCfgSTT(is_cuda=False, uuid='db3cb3cd9b', cache_folder='/Users/xiaohuo/pyvideotrans-main/tmp/8951/db3cb3cd9b', target

219.255* Posted at: 1 day ago

语音识别阶段出错 [faster-whisper(本地)] Unknow error
Traceback (most recent call last):
File "/Users/xiaohuo/pyvideotrans-main/videotrans/task/job.py", line 105, in run

trk.recogn()

File "/Users/xiaohuo/pyvideotrans-main/videotrans/task/_speech2text.py", line 152, in recogn

raw_subtitles = run(

File "/Users/xiaohuo/pyvideotrans-main/videotrans/recognition/__init__.py", line 265, in run

return FasterAll(**kwargs).run()

File "/Users/xiaohuo/pyvideotrans-main/videotrans/recognition/_base.py", line 145, in run

raise RuntimeError('Unknow error')

RuntimeError: Unknow error
TaskCfgSTT(is_cuda=False, uuid='db3cb3cd9b', cache_folder='/Users/xiaohuo/pyvideotrans-main/tmp/8951/db3cb3cd9b', target_dir='/Users/xiaohuo/Downloads/pyvideotrans-main/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='/Users/xiaohuo/Downloads/pyvideotrans-main/output/recogn/耶稣教唱歌.srt', target_wav=None, target_wav_output=None, name='/Users/xiaohuo/Downloads/耶稣教唱歌.m4a', noextname='耶稣教唱歌', basename='耶稣教唱歌.m4a', ext='m4a', dirname='/Users/xiaohuo/Downloads', shound_del_name=None, detect_language='ko', recogn_type=0, model_name='medium', shibie_audio='/Users/xiaohuo/pyvideotrans-main/tmp/8951/db3cb3cd9b/耶稣教唱歌-1781426490.165096.wav', remove_noise=False, enable_diariz=False, nums_diariz=1, rephrase=0, fix_punc=False)


=
system:macOS-26.5.1-arm64-arm-64bit
version:v3.98
frozen:False
language:zh
root_dir:/Users/xiaohuo/pyvideotrans-main
Python: 3.10.19 (main, Feb 12 2026, 00:36:33) [Clang 21.1.4 ]

1 Replies

语音识别阶段报 Unknow error,通常由以下三种原因之一导致,按可能性排列:

  1. 模型下载不完整或路径错误faster-whispermedium 模型需事先下载到 models/ 目录。请检查 /Users/xiaohuo/pyvideotrans-main/models/ 下是否存在 models--Systran--faster-whisper-medium 文件夹且包含 .bin 等文件。若缺失,请手动下载(参见 faster-whisper模型下载说明)。
  2. 中文路径/文件名:输入文件 耶稣教唱歌.m4a 包含中文且位于中文目录。源码部署环境下,底层C库对非ASCII路径兼容性差,可能引发静默崩溃。请将文件重命名为纯英文(如 song.m4a)并移至纯英文路径,再重试。
  3. 内存不足medium 模型在CPU模式下需约8GB以上可用内存。macOS ARM 设备若内存不足(如8GB Mac),模型加载时可能静默失败。建议改用 basetiny 模型测试。

诊断步骤:

  1. 关闭软件,将文件改名为 test.m4a 并复制到 /Users/xiaohuo/Downloads/ 目录。
  2. 打开软件,进入 批量语音转字幕 功能,选择该文件,ASR渠道选 faster-whisper(本地),模型选 base,取消勾选CUDA(你的设备不支持),点击执行。
  3. 若仍报错,请提供 logs/ 文件夹中最新 .log 文件末尾30行内容。

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. Dondate: ko-fi.com/jianchang512

Related resource