OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-07 12:21:27 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity
f89ae9cb8d
pytorch/docs
History
Michael Carilli f89ae9cb8d Moves grid_sampler to autocast promote list (#58618) 
Summary:
Should close https://github.com/pytorch/pytorch/issues/42218

Numerically, `grid_sampler` is fine in fp16 or fp32, but takes several inputs and expects their dtypes to match, so it belongs on the autocast promote list.

`grid_sampler` currently uses `gpuAtomicAdd`, notoriously slow in fp16 because it calls cuda's atomicAdd __half overload which uses a software compare-and-swap loop internally. To allow good performance if both inputs happen to be FP16, the PR also modifies `grid_sampler_[2,3]d_backward_kernel`s to use `fastAtomicAdd` instead.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/58618

Reviewed By: mruberry

Differential Revision: D29257199

Pulled By: ngimel

fbshipit-source-id: 3cc7505945b480427f2fc1beb36bee80bf3853b3
2021-06-21 10:22:36 -07:00
..
caffe2 Lint trailing newlines (#54737) 2021-03-30 13:09:52 -07:00
cpp Add no-grad inference mode note (#58513) 2021-05-25 13:06:54 -07:00
source Moves grid_sampler to autocast promote list (#58618) 2021-06-21 10:22:36 -07:00
.gitignore .gitignore for the docs folder 2019-10-08 12:18:30 -07:00
libtorch.rst DOC: Building libtorch using CMake (#44196) 2020-10-21 14:29:36 -07:00
make.bat Sphinx parallel build (#38785) 2020-05-21 13:03:55 -07:00
Makefile DOC: fail to build if there are warnings (#41335) 2020-07-28 22:33:44 -07:00
README.md Add docs/README.md to make existing doc build info more discoverable (#49286) 2020-12-16 11:55:45 -08:00
requirements.txt [1/n][torch/elastic] Move torchelastic docs *.rst (#148) 2021-05-04 00:57:56 -07:00

README.md

Please see the Writing documentation section of CONTRIBUTING.md for details on both writing and building the docs.