#1496 在使用人声背景声分离时报错Traceback (most recent call last): File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24

这个错误表明人声背景声分离功能缺少必要的模型文件。错误信息显示无法找到 uvr5_weights/modelparams/2band_44100_lofi.json 文件。

解决方案：

1. 重新下载完整软件包：您可能使用的是不完整的版本。请前往官方网站 https://pyvideotrans.com/downpackage 下载完整的 2.5GB 软件包，解压到纯英文路径下使用。

2. 手动下载模型文件：如果重新下载后问题依旧，可以尝试手动下载 UVR5 模型：

   • 访问 HuggingFace 仓库下载相关模型文件
   • 将下载的 uvr5_weights 文件夹放置到软件根目录下
3. 检查软件完整性：确保软件目录结构完整，特别是 uvr5_weights/modelparams/ 目录及其中的 JSON 配置文件。
4. 临时解决方案：如果急需处理视频，可以暂时不勾选"保留原始背景音"选项，这样会跳过人声分离步骤。

如果以上方法无法解决问题，请提供您使用的软件版本信息。

我是window用uv源码构建的,如何修复

您在使用源码构建时缺少 UVR5 模型文件。请按以下步骤解决：

1. 手动下载 UVR5 模型文件：

   • 访问 HuggingFace 仓库：https://huggingface.co/TRvlvr/model_repo/resolve/main/uvr5_weights.zip
   • 下载完整的 uvr5_weights.zip 文件

2. 解压并放置文件：

   • 将下载的 uvr5_weights.zip 解压
   • 将解压后的 uvr5_weights 文件夹放置到您的源码目录 D:\MyGithubClone\pyvideotrans\ 下
   • 确保文件路径为：D:\MyGithubClone\pyvideotrans\uvr5_weights\modelparams\2band_44100_lofi.json

3. 验证文件结构：

   • 检查 uvr5_weights 文件夹应包含：

     • modelparams/ 目录（包含各种 JSON 配置文件）
     • onnx/ 目录（包含模型文件）

4. 重新运行软件：

   • 完成文件放置后，重新启动 pyVideoTrans 即可

如果下载困难，也可以从官方发布的预编译版本中提取 uvr5_weights 文件夹。

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = f

......
ithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

您下载的 UVR5 模型文件不完整，缺少关键的模型权重文件 HP2.pth。请按以下步骤解决：

1. 下载完整模型包：

   • 访问：https://huggingface.co/TRvlvr/model_repo/resolve/main/uvr5_weights.zip
   • 下载完整的 uvr5_weights.zip 文件（约 1.5GB）

2. 检查文件完整性：

   • 解压后确保 uvr5_weights 文件夹包含：

     • modelparams/ 目录（JSON 配置文件）
     • onnx/ 目录（ONNX 模型文件）
     • HP2.pth、HP5.pth 等 .pth 权重文件

3. 替换文件：

   • 将下载的完整 uvr5_weights 文件夹替换您当前不完整的版本
   • 确保路径为：D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth

4. 验证：

   • 检查 HP2.pth 文件是否存在且文件大小正常（约几百MB）

如果下载困难或文件损坏，建议从官方预编译版本中提取完整的 uvr5_weights 文件夹。

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = fun

......
r, mode)
File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

Traceback (most recent call last):
File "D:\MyGithubClone\pyvideotrans\videotrans\separate\st.py", line 24, in uvr

pre_fun = func(

File "D:\MyGithubClone\pyvideotrans\videotrans\separate\vr.py", line 32, in init

cpk = torch.load(model_path, map_location="cpu")

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 1479, in load

with _open_file_like(f, "rb") as opened_file:

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 759, in _open_file_like

return _open_file(name_or_buffer, mode)

File "D:\MyGithubClone\pyvideotrans.venv\lib\site-packages\torch\serialization.py", line 740, in init

super().__init__(open(name, mode))

FileNotFoundError: [Errno 2] No such file or directory: 'D:/MyGithubClone/pyvideotrans/uvr5_weights/HP2.pth'

等待 uuid_signal 线程退出
Attempting to kill ffmpeg processes for user: PC
Successfully killed ffmpeg processes using taskkill
请按任意键继续. . .

您下载的 UVR5 模型文件仍然不完整，缺少 HP2.pth 权重文件。请按以下步骤彻底解决：

1. 下载完整模型包：

   • 访问：https://huggingface.co/TRvlvr/model_repo/resolve/main/uvr5_weights.zip
   • 确保下载完整的 1.5GB 文件，不要使用不完整的下载

2. 验证文件完整性：

   • 解压后检查 uvr5_weights 文件夹应包含：

     • HP2.pth（约 170MB）
     • HP5.pth（约 170MB）
     • 其他 .pth 文件
     • modelparams/ 目录
     • onnx/ 目录

3. 替代方案：

   • 如果 HuggingFace 下载困难，从官方预编译版本中提取：

     • 下载官方 2.5GB 完整包
     • 从中复制完整的 uvr5_weights 文件夹到您的源码目录

4. 临时解决方案：

   • 如果急需处理视频，暂时不勾选"保留原始背景音"选项，跳过人声分离步骤

请确保 HP2.pth 文件确实存在于 D:/MyGithubClone/pyvideotrans/uvr5_weights/ 目录中。

我成功了
再问下,我分离背景声后, 重新配音又要重新分离, 如何逃过此步骤

pyVideoTrans 有缓存机制可以避免重复分离背景声。请按以下步骤操作：

1. 不要勾选"清理已生成"：在主界面确保"清理已生成"复选框未勾选，这样软件会使用之前分离的缓存文件。

2. 检查缓存位置：分离后的背景音文件保存在：

   • 软件根目录下的 tmp 文件夹
   • 或用户目录下的 pyvideotrans 文件夹

3. 重新处理时：

   • 保持相同的视频文件路径
   • 使用相同的分离设置
   • 软件会自动检测并使用已有的背景音文件
4. 如果缓存失效：可能是视频文件路径改变或软件重启导致缓存丢失。此时需要重新分离一次，但后续处理会继续使用缓存。

注意：只要不手动清理缓存或勾选"清理已生成"，背景音分离结果会被重复使用。