Open malikeh1375 opened 3 years ago
same question, any solution?
In pytorch you can call the .to() function to attach a device to another GPU / CPU.
tensor.to('cpu') should attach the tensor to the CPU of your computer. However, I believe that quite a lot of tensors should be reattched so I would recommend that you be careful with this.
You can set CUDA = int(False) in the run_[model].py.
I am running the codes on macOS without NVIDIA GPU Graphic Card available. I wonder how I can run the scripts in CPU-only mode. What sections should I modify?
Thanks in advance!