File: nn.init.rst

package info (click to toggle)
pytorch 2.9.1%2Bdfsg-1~exp2
  • links: PTS, VCS
  • area: main
  • in suites: experimental
  • size: 180,096 kB
  • sloc: python: 1,473,255; cpp: 942,030; ansic: 79,796; asm: 7,754; javascript: 2,502; java: 1,962; sh: 1,809; makefile: 628; xml: 8
file content (29 lines) | stat: -rw-r--r-- 778 bytes parent folder | download | duplicates (5)
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
.. role:: hidden
    :class: hidden-section

.. _nn-init-doc:

torch.nn.init
=============

.. warning::
    All the functions in this module are intended to be used to initialize neural network
    parameters, so they all run in :func:`torch.no_grad` mode and will not be taken into
    account by autograd.

.. currentmodule:: torch.nn.init
.. autofunction:: calculate_gain
.. autofunction:: uniform_
.. autofunction:: normal_
.. autofunction:: constant_
.. autofunction:: ones_
.. autofunction:: zeros_
.. autofunction:: eye_
.. autofunction:: dirac_
.. autofunction:: xavier_uniform_
.. autofunction:: xavier_normal_
.. autofunction:: kaiming_uniform_
.. autofunction:: kaiming_normal_
.. autofunction:: trunc_normal_
.. autofunction:: orthogonal_
.. autofunction:: sparse_