projects / platform / upstream / caffe.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: c5e404d) | patch
cuDNN pooling layer tests know that nonzero padding is not supported
authorJonathan L Long <jonlong@cs.berkeley.edu>
Tue, 30 Sep 2014 04:21:03 +0000 (21:21 -0700)
committerJonathan L Long <jonlong@cs.berkeley.edu>
Tue, 30 Sep 2014 04:25:06 +0000 (21:25 -0700)
commit618b92934b4d1449c10b6611c87d56be9a4d75d2
tree61f623b4a1d346ddb27e595db7158922a21f0bd0tree | snapshot
parentc5e404dc2988532906b08eb90b226be91b778ac0commit | diff
cuDNN pooling layer tests know that nonzero padding is not supported

After #1172, some cuDNN pooling layer tests were failing due to use of
padding. (Before #1172, these tests were actually testing PoolingLayer
instead of CuDNNPoolingLayer via the fallback.) This commit disables
many tests via commenting, so that they can be easily readded when cuDNN
gains pooling padding support.
src/caffe/test/test_pooling_layer.cpp diff | blob | history
Domain: Machine Learning / ML Framework;