4 Commits

Author	SHA1	Message	Date
Lin Li	4a2f3cc45f	Improve lars operator by applying clipping (#9905) ... Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/9905 This diff improves lars operator in Caffe2 by applying clipping to the computed learning rate Reviewed By: pjh5 Differential Revision: D9020606 fbshipit-source-id: b579f1d628113c09366feac9406002f1ef4bd54f	2018-08-02 11:54:28 -07:00
Orion Reblitz-Richardson	1d5780d42c	Remove Apache headers from source. ... * LICENSE file contains details, so removing from individual source files.	2018-03-27 13:10:18 -07:00
Qinqing Zheng	d013e16cf4	[C2] Enable LARS on GPU (#2115)	2018-03-02 18:06:19 -08:00
Qinqing Zheng	7cafdab69b	[C2] Implement Layer-wise Adaptive Rate Scaling (LARS) (#2034) ... * [C2] Implement Layer-wise Adaptive Rate Scaling (LARS) * [C2] Implement Layer-wise Adaptive Rate Scaling (LARS) * add unit test for Lars * set default value for lars to be None * remove lars for subclasses of SgdOptimizer	2018-02-25 14:58:31 -08:00