Dealing with a couple errors when running without GPU. One being print(torch.cuda.get_device_name(0)) is called even when installed without cuda, the other checkpoint = torch.load('trained.torch_model') mapped to GPU.
Also removing requiresgrad() as per Georg's comment.
Dealing with a couple errors when running without GPU. One being
print(torch.cuda.get_device_name(0))
is called even when installed without cuda, the othercheckpoint = torch.load('trained.torch_model')
mapped to GPU.Also removing requiresgrad() as per Georg's comment.