34j comments

Results 171 comments of

34j

MPS Support

I believe there was a feature to do that automatically (in lightning?)

MPS Support

It seems that environmental variable PYTORCH_ENABLE_MPS_FALLBACK should be set to 1, please try that https://stackoverflow.com/a/72416727

MPS Support

Try replacing here with if True: to patch stft if isinstance(self.trainer.accelerator, TPUAccelerator): https://github.com/34j/so-vits-svc-fork/blob/main/src/so_vits_svc_fork/train.py#L179-L179

MPS Support

I mean ```python if True: # patch torch.stft to use cpu LOG.warning("Using TPU. Patching torch.stft to use cpu.") ```

MPS Support

@allcontributors add escoolioinglesias bug, userTesting

RuntimeError: Expected 2D (unbatched) or 3D (batched) input to conv1d, but got input of size: [6, 2, 80, 1649]

It might be because your dataset contains stereo audio files or maybe not.

Specifying `args` in `pre-commit-config.yaml` causes no formatting (and no output)

You are right, I tried it and confirmed that it does not work with 1.10.0 and works with 1.9.0. I have also tried the CLI, and the following commands seem...

Error when running multiple CUDA GPUs

You can just call tensorboard manually ``` tensorboard --logdir logs/44k ```

svcg did not work!

https://github.com/PySimpleGUI/PySimpleGUI/issues/4703

Results are very noisy? (Mac M1 Max)

Is it simply because the device of the model is not changed to CPU when saved? If so, I think #1149 will help