Search results

Why the BN layer of your deploy.prototxt does not set "use_global_stats: true"

Status: Open.

#15 In farmingyard/caffe-mobilenet;

· Ai-is-light opened on Mar 7, 2018

depthwise layer is the same with original group convolution？

Status: Open.

#13 In farmingyard/caffe-mobilenet;

· lbwang2006 opened on Dec 27, 2017

"caffe_gpu_set" should be "caffe_set" in line 125 and 139

Status: Open.

#12 In farmingyard/caffe-mobilenet;

· jnulzl opened on Dec 23, 2017

Loss diverges on using ConvolutionDepthwise layer.

Status: Open.

#10 In farmingyard/caffe-mobilenet;

· krraush opened on Oct 27, 2017

How this solution do group operation?I don't understand ,pls help me,thanks

Status: Open.

#9 In farmingyard/caffe-mobilenet;

· lyk125 opened on Sep 22, 2017

shoul i add your ConvolutionDepthwise layer to the caffe.proto？

Status: Open.

#6 In farmingyard/caffe-mobilenet;

· 123chengbo opened on Aug 25, 2017

caffe crash after build with this new kernel

Status: Open.

#5 In farmingyard/caffe-mobilenet;

· ouceduxzk opened on Aug 24, 2017

Training Accuracy problem

Status: Open.

#4 In farmingyard/caffe-mobilenet;

· leochli opened on Aug 16, 2017

how about the speed of depth convolution on cpu VS original caffe?

Status: Open.

#2 In farmingyard/caffe-mobilenet;

· lyk125 opened on Aug 3, 2017

hpp/cpp/cu files compile error

Status: Open.

#1 In farmingyard/caffe-mobilenet;

· leochli opened on Jul 19, 2017