cpu or gpu

[A-Rain]

When training the model, can we use CPU instead of GPU?

[gvanhorn38]

The easiest way to do this would be to set the environment variableCUDA_VISIBLE_DEVICES=3, where 3 is a number that is larger than the number of gpus present in your machine. This will prevent tensorflow from using the gpus. So you can do something like:

$ CUDA_VISIBLE_DEVICES=3 python train.py ....