Whether this whl package can be used to run the model on GPU.
True if CUDA is currently available, otherwise False.
Bool
Examples
>>> import paddle >>> support_gpu = paddle.device.is_compiled_with_cuda()