projects / platform / upstream / pytorch.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: e3da16a) | patch
Optimize MomentumSGDUpdate maximum block size and make it templated
authorBilge Acun <acun@fb.com>
Fri, 22 Mar 2019 16:51:27 +0000 (09:51 -0700)
committerFacebook Github Bot <facebook-github-bot@users.noreply.github.com>
Fri, 22 Mar 2019 16:54:25 +0000 (09:54 -0700)
commitf6df6aed89c00f3baa270998417ce2b8ca5756c9
tree613881a10f79cd904c809b1fd98a2acda41a72adtree | snapshot
parente3da16a99eb4b475a90fe424eaf9156c3f7370f4commit | diff
Optimize MomentumSGDUpdate maximum block size and make it templated

Summary: Removing the maximum number of blocks limit from the operator and making the nesterov parameter templated to remove branching.

Reviewed By: BIT-silence

Differential Revision: D14567003

fbshipit-source-id: 394c2039ee214adc6ccd2e562e4e9563d307131f
caffe2/sgd/momentum_sgd_op_gpu.cu diff | blob | history
Domain: Machine Learning / ML Framework;