projects / platform / upstream / pytorch.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: afc7574) | patch
add fbgemm fp16 (fbfcpacked) support, add global_init_net in predictor_export_meta...
authorWeiyi Zheng <wyz@fb.com>
Fri, 22 Mar 2019 07:08:50 +0000 (00:08 -0700)
committerFacebook Github Bot <facebook-github-bot@users.noreply.github.com>
Fri, 22 Mar 2019 07:19:59 +0000 (00:19 -0700)
commitf3cf6ed789683db0f2e1a5da9926b82f935d9d3b
tree7dcaf6c37d04892f2221382168fd96b67b372a74tree | snapshot
parentafc7574aed08a1a37ec132489a7010eed04da66fcommit | diff
add fbgemm fp16 (fbfcpacked) support, add global_init_net in predictor_export_meta (#18257)

Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/18257

support adding op in global_init_net. because pred_init_net is per thread, and just doesn't cut it.

Reviewed By: jspark1105

Differential Revision: D14552695

fbshipit-source-id: 53dd44c84ad019019ab9f35fc04d076b7f941ddc
caffe2/python/predictor/predictor_exporter.py diff | blob | history
caffe2/python/predictor/predictor_exporter_test.py diff | blob | history
Domain: Machine Learning / ML Framework;