1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
|
```{eval-rst}
.. currentmodule:: torch
```
(type-info-doc)=
# Type Info
The numerical properties of a {class}`torch.dtype` can be accessed through either the {class}`torch.finfo` or the {class}`torch.iinfo`.
(finfo-doc)=
## torch.finfo
```{eval-rst}
.. class:: torch.finfo
```
A {class}`torch.finfo` is an object that represents the numerical properties of a floating point
{class}`torch.dtype`, (i.e. ``torch.float32``, ``torch.float64``, ``torch.float16``, and ``torch.bfloat16``).
This is similar to [numpy.finfo](https://numpy.org/doc/stable/reference/generated/numpy.finfo.html).
A {class}`torch.finfo` provides the following attributes:
| Name | Type | Description |
| :-------------- | :---- | :------------------------------------------------------------------------------------------ |
| bits | int | The number of bits occupied by the type. |
| eps | float | The difference between 1.0 and the next smallest representable float larger than 1.0. |
| max | float | The largest representable number. |
| min | float | The smallest representable number (typically ``-max``). |
| tiny | float | The smallest positive normal number. Equivalent to ``smallest_normal``. |
| smallest_normal | float | The smallest positive normal number. See notes. |
| resolution | float | The approximate decimal resolution of this type, i.e., ``10**-precision``. |
```{note}
The constructor of {class}`torch.finfo` can be called without argument,
in which case the class is created for the pytorch default dtype (as returned by {func}`torch.get_default_dtype`).
```
```{note}
`smallest_normal` returns the smallest *normal* number, but there are smaller
subnormal numbers. See https://en.wikipedia.org/wiki/Denormal_number
for more information.
```
(iinfo-doc)=
## torch.iinfo
```{eval-rst}
.. class:: torch.iinfo
```
A {class}`torch.iinfo` is an object that represents the numerical properties of a integer
{class}`torch.dtype` (i.e. ``torch.uint8``, ``torch.int8``, ``torch.int16``, ``torch.int32``, and ``torch.int64``).
This is similar to [numpy.iinfo](https://numpy.org/doc/stable/reference/generated/numpy.iinfo.html).
A {class}`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. |
|
