File: cudnn_persistent_rnn.rst

package info (click to toggle)
pytorch 1.7.1-7
  • links: PTS, VCS
  • area: main
  • in suites: bullseye
  • size: 80,340 kB
  • sloc: cpp: 670,830; python: 343,991; ansic: 67,845; asm: 5,503; sh: 2,924; java: 2,888; xml: 266; makefile: 244; ruby: 148; yacc: 144; objc: 51; lex: 44
file content (12 lines) | stat: -rw-r--r-- 321 bytes parent folder | download 
1
2
3
4
5
6
7
8
9
10
11
12
 
:orphan:


.. note::

    If the following conditions are satisfied:
    1) cudnn is enabled, 
    2) input data is on the GPU 
    3) input data has dtype ``torch.float16`` 
    4) V100 GPU is used,
    5) input data is not in ``PackedSequence`` format
    persistent algorithm can be selected to improve performance.