torch.set_default_dtype

torch.set_default_dtype(d, /)

将默认浮点数据类型设置为 d。支持浮点数据类型作为输入。其他数据类型将导致torch引发异常。

当PyTorch初始化时，其默认浮点数据类型为torch.float32，而set_default_dtype(torch.float64)的目的是为了便于NumPy风格的类型推断。默认浮点数据类型用于

1. 隐式确定默认复数数据类型。当默认浮点类型为float16时，默认复数数据类型为complex32。对于float32，默认复数数据类型为complex64。对于float64，它是complex128。对于bfloat16，将引发异常，因为bfloat16没有对应的复数类型。

2. 推断使用Python浮点数或复数构建的张量的数据类型。请参阅下面的示例。

3. 确定布尔值和整数张量与Python浮点数和复数之间的类型提升结果。

参数

d (torch.dtype) – 要设置为默认值的浮点数据类型。

示例

>>> # initial default for floating point is torch.float32
>>> # Python floats are interpreted as float32
>>> torch.tensor([1.2, 3]).dtype
torch.float32
>>> # initial default for floating point is torch.complex64
>>> # Complex Python numbers are interpreted as complex64
>>> torch.tensor([1.2, 3j]).dtype
torch.complex64

>>> torch.set_default_dtype(torch.float64)
>>> # Python floats are now interpreted as float64
>>> torch.tensor([1.2, 3]).dtype  # a new floating point tensor
torch.float64
>>> # Complex Python numbers are now interpreted as complex128
>>> torch.tensor([1.2, 3j]).dtype  # a new complex tensor
torch.complex128

>>> torch.set_default_dtype(torch.float16)
>>> # Python floats are now interpreted as float16
>>> torch.tensor([1.2, 3]).dtype  # a new floating point tensor
torch.float16
>>> # Complex Python numbers are now interpreted as complex128
>>> torch.tensor([1.2, 3j]).dtype  # a new complex tensor
torch.complex32