#7564 Subprocess Error: Error no file named model.safetensors, or pytorch_model.bin, found in directory F:/sppy/models/models-

183.167**3 Posted at: 1 hour ago

Subprocess Error: Error no file named model.safetensors, or pytorch_model.bin, found in directory F:/sppy/models/models--k2-fsa--OmniVoice.:
{'text': '你好啊,我亲爱的朋友,希望你的每一天都是美好愉快的!', 'rate': '+0%', 'role': '쏟아져 나오는 중국의 로봇! [코너별 다시보기 - 0921].ae-0-5.4.wav', 'filename': 'C:\Users\windos\AppData\Local\Temp/pyvideotrans/3-1790087448.8593526.wav', 'tts_type': 3, 'language': 'zh-cn', 'volume': '+0%', 'pitch': '+0Hz'}
Traceback (most recent call last):
File "videotrans\util\ListenVoice.py", line 21, in run
File "videotrans\tts\__init__.py", line 232, in run
File "videotrans\tts\_base.py", line 97, in run
File "videotrans\tts\_omnivoice.py", line 42, in _exec
File "videotrans\configure\base.py", line 264, in _new_process
videotrans.configure.excepts.VideoTransError: Subprocess Error: Error no file named model.safetensors, or pytorch_model.bin, found in directory F:/sppy/models/models--k2-fsa--OmniVoice.


=
system:Windows-10-10.0.22631-SP0
version:v4.12
frozen:True
language:zh_CN
root_dir:F:/sppy

1 Replies

#1 DeepSeek reply at 1 hour ago

核心原因是 OmniVoice 模型文件缺失或不完整:目录 F:/sppy/models/models--k2-fsa--OmniVoice 下没有 model.safetensorspytorch_model.bin,说明该内置模型未成功下载(或只下载了部分文件)。

模型文件普遍较大(1G-5G 以上),国内直连 huggingface.co 常失败,自动下载中断后目录里可能残留不完整文件,从而触发此报错。

解决方案:

  1. 彻底删除损坏的模型目录:
    删除 F:/sppy/models/models--k2-fsa--OmniVoice 整个文件夹。
  2. 手动下载 OmniVoice 模型全部文件:
    下载地址:https://huggingface.co/k2-fsa/OmniVoice/tree/main
  3. 下载完成后放到:
    F:/sppy/models/models--k2-fsa--OmniVoice

    注意:

    • 必须下载该页面下的全部文件(包括 model.safetensors、配置文件等),缺一不可。
    • 若浏览器下载时自动重命名为 model(1).safetensors 之类,必须改回原名 model.safetensors,否则软件无法识别。
    • 下载前建议确认已科学上网;若不使用代理,软件内部会自动切换 hf-mirror.com 镜像,但镜像有频率限制,仍可能失败。
  4. 放好后重新启动软件,再次选中 OmniVoice 渠道配音测试。
附带提示:你的 role(参考音频角色名)是韩文视频名 쏟아져 나오는 중국의 로봇!...wav,说明你用的是 clone 克隆音色,需确保参考音频时长在 3-10 秒、且高级选项中「最短语音持续」设为 3000ms 以上,否则克隆同样容易失败。

请查阅相关文档:

Thinking...
Post Your Reply
If you have more than 50 messages, you will no longer receive an automatic reply. Please ask a new question!
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 via ko-fi.com

Related resource