【pytorch】数据类型及转换方式

写在前面

张量是一个包含单一数据类型元素的多维矩阵，pytorch中共定义了10种张量的数据类型。

数据类型

Data type	dtype	CPU tensor	GPU tensor
32-bit floating point	torch.float32 or torch.float	torch.FloatTensor	torch.cuda.FloatTensor
64-bit floating point	torch.float64 or torch.double	torch.DoubleTensor	torch.cuda.DoubleTensor
16-bit floating point 1	torch.float16 or torch.half	torch.HalfTensor	torch.cuda.HalfTensor
16-bit floating point 2	torch.bfloat16	torch.BFloat16Tensor	torch.cuda.BFloat16Tensor
32-bit complex	torch.complex32	\	\
64-bit complex	torch.complex64	\	\
128-bit complex	torch.complex128 or torch.cdouble	\	\
8-bit integer (unsigned)	torch.uint8	torch.ByteTensor	torch.cuda.ByteTensor
8-bit integer (signed)	torch.int8	torch.CharTensor	torch.cuda.CharTensor
16-bit integer (signed)	torch.int16 or torch.short	torch.ShortTensor	torch.cuda.ShortTensor
32-bit integer (signed)	torch.int32 or torch.int	torch.IntTensor	torch.cuda.IntTensor
64-bit integer (signed)	torch.int64 or torch.long	torch.LongTensor	torch.cuda.LongTensor
Boolean	torch.bool	torch.BoolTensor	torch.cuda.BoolTensor

转换方式

方式1

直接在张量后面加.int()、.long()、.float()、.double()

>>> import torch
>>> a = torch.tensor([1.3, 1.5, 1.7])
>>> a
tensor([1.3000, 1.5000, 1.7000])
>>> a.dtype
torch.float32
>>> a.int()
tensor([1, 1, 1], dtype=torch.int32)
>>> a.long()
tensor([1, 1, 1])
>>> a.float()
tensor([1.3000, 1.5000, 1.7000])
>>> a.double()
tensor([1.3000, 1.5000, 1.7000], dtype=torch.float64)

方式2

使用张量的.type(dtype)方法，dtype参数填数据类型表格中的第二列，如torch.uint8

>>> import torch
>>> a = torch.tensor([-1, 1, 256])
>>> a
tensor([ -1,   1, 256])
>>> a.dtype
torch.int64
>>> a.type(torch.uint8)
tensor([255,   1,   0], dtype=torch.uint8)

方式3

使用张量的.type_as(tensor)方法将张量1转换为和张量2一样的数据类型

>>> import torch
>>> a = torch.tensor([1, 2, 3])
>>> a.dtype
torch.int64
>>> b = torch.tensor([1.1, 2.2, 3.3])
>>> b.dtype
torch.float32
>>> b = b.type_as(a)
>>> b.dtype
torch.int64
>>> b
tensor([1, 2, 3])

引用参考

https://pytorch.org/docs/stable/tensors.html