Minor inconsistency in 3.2.1 - [ CPUFloatType{5} ] should read [ CUDALongType{5} ]

[FvD]

Thank you for writing and sharing this great book.

I noticed a minor inconsistency in 3.2.1 Tensors from values. Where it currently reads:

Analogously, the default device is the CPU; but we can also create a tensor that, right from the outset, is located on the GPU:

torch_tensor(1:5, device = "cuda")

torch_tensor
1
2
3
4
5
[ CPUFloatType{5} ]

the output should read: [ CUDALongType{5} ].