The numerical properties of a torch.dtype
can be accessed through either the torch.finfo
or the torch.iinfo
.
class torch.finfo
A torch.finfo
is an object that represents the numerical properties of a floating point torch.dtype
, (i.e. torch.float32
, torch.float64
, and torch.float16
). This is similar to numpy.finfo.
A torch.finfo
provides the following attributes:
Name | Type | Description |
---|---|---|
bits | int | The number of bits occupied by the type. |
eps | float | The smallest representable number such that |
max | float | The largest representable number. |
min | float | The smallest representable number (typically |
tiny | float | The smallest positive representable number. |
resolution | float | The approximate decimal resolution of this type, i.e., |
Note
The constructor of torch.finfo
can be called without argument, in which case the class is created for the pytorch default dtype (as returned by torch.get_default_dtype()
).
class torch.iinfo
A torch.iinfo
is an object that represents the numerical properties of a integer torch.dtype
(i.e. torch.uint8
, torch.int8
, torch.int16
, torch.int32
, and torch.int64
). This is similar to numpy.iinfo.
A torch.iinfo
provides the following attributes:
Name | Type | Description |
---|---|---|
bits | int | The number of bits occupied by the type. |
max | int | The largest representable number. |
min | int | The smallest representable number. |
© 2019 Torch Contributors
Licensed under the 3-clause BSD License.
https://pytorch.org/docs/1.7.0/type_info.html