RuntimeError: Failed to load checkpoint

[sonhm3029]

Describe the bug

Traceback (most recent call last):
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/so_vits_svc_fork/train.py", line 347, in load
    _, _, _, epoch = utils.load_checkpoint(
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/so_vits_svc_fork/utils.py", line 264, in load_checkpoint
    checkpoint_dict = torch.load(f, map_location="cpu", weights_only=False)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/torch/serialization.py", line 712, in load
    return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/torch/serialization.py", line 1047, in _load
    unpickler = UnpicklerWrapper(data_file, **pickle_load_args)
TypeError: __init__() got an unexpected keyword argument 'weights_only'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/home/edabk/anaconda3/envs/sovits_3_9/bin/svc", line 8, in <module>
    sys.exit(cli())
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/click/core.py", line 1130, in __call__
    return self.main(*args, **kwargs)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/click/core.py", line 1055, in main
    rv = self.invoke(ctx)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/click/core.py", line 1657, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/click/core.py", line 1404, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/click/core.py", line 760, in invoke
    return __callback(*args, **kwargs)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/so_vits_svc_fork/__main__.py", line 128, in train
    train(
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/so_vits_svc_fork/train.py", line 119, in train
    model = VitsLightning(reset_optimizer=reset_optimizer, **hparams)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/so_vits_svc_fork/train.py", line 186, in __init__
    self.load(reset_optimizer)
  File "/home/edabk/anaconda3/envs/sovits_3_9/lib/python3.9/site-packages/so_vits_svc_fork/train.py", line 363, in load
    raise RuntimeError("Failed to load checkpoint") from e
RuntimeError: Failed to load checkpoint

I receive this error while running python=3.9.13 with torch version 1.12.1 and cuda-toolkit 11.3.

It seem like my pytorch version not have weight_only param in torch.load.

I've use torch+cu118 like in the guideline but receive the error that No computation device found ( I think that because of my cuda version not matching) so i 've to use torch=1.12.1 with cuda version 11.3

To Reproduce

conda create -n sovits python=3.9.13
conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.3 -c pytorch
pip install so-vits-svc-fork

train svc

Additional context

No response

Version

lastest

Platform

Ubuntu 20.04

Code of Conduct

• I agree to follow this project's Code of Conduct.

No Duplicate

• I have checked existing issues to avoid duplicates.

[sbersier]

The weights_only parameter was indeed added in torch==1.13

Personally, I would install it like this:

mkdir MySovitsFolder
cd MySovitsFolder
python -m venv venv
source venv/bin/activate  # to activate the venv environnment
pip install -U torch torchaudio

In principle, it should install torch==2.0 with everything needed.
You can check that everything is fine with:
python -c "import torch; print(torch.__version__)"

Then:

pip install -U so-vits-svc-fork
svcg

And:
deactivate # to deactivate the environment