How to run this on CPU device

[rayguang]

I am running a Macbook M1 and it seems to take quite some work to work with the device mps.

I am checking if there is an easy way to just run the code on cpu device. I have tried to set --gpu_ids to -1 when running for example test.py but no luck,

I used the following versions of torch and python 3.9 torch 1.13.1 via pypi

[Ryukhaan]

I have the same problem.

I tried to set "mps" instead of "cuda" in test.py - line 39 device = torch.device("mps" if torch.backends.mps.is_available() else rank)

However, several erros still occur, mainly in Nvdiffrast library.