Skip to first unread message
Pankaj Randhe
Apr 21, 2016, 2:27:16 PM4/21/16
to Caffe Users
For some reasons, I am using Caffe rc2 . While using the trained model for the prediction I am getting the following error which doesn't appear if I use rc3 of Caffe. My data blob size is 50*20*5*5.
F0101 07:14:20.486869 11860 net.cpp:710] Check failed: target_blobs[j]->num() == source_layer.blobs(j).num() (20 vs. 0)
What could be the reason behind this error and how can I rectify this. I am compelled to use rc2 of Caffe because of CuDNN v2 is the highest supported version on Jetson TK1 due to its CUDA 6.5 version.
Thank You.
Reply all
Reply to author
Forward
0 new messages