some tests about torch tensor

if you want to transform your tensor to other types,you may do like this

after all,you need to make sure the type of the tensor’s type

import torch

tensor=torch.rand(2,1)
print(tensor.dtype())

import torch

tensor=torch.rand(2,1)
print(tensor)
print(tensor.dtype)
#tensor([[0.0863],
#        [0.6352]])
#torch.float32


tensor=tensor.long()
print(tensor.dtype)
#torch.int64

tensor=tensor.int()
print(tensor.dtype)
#torch.int32

tensor=tensor.float()
print(tensor.dtype)
#torch.float32

tensor=tensor.double()
print(tensor.dtype)
#torch.float64

tensor=tensor.half()
print(tensor.dtype)
#torch.float16

tensor=tensor.byte()
print(tensor.dtype)
#torch.uint8

tensor=tensor.char()
print(tensor.dtype)
#torch.int8

the results is like this