The training hangs for solver_cfn_jps.prototxt

[jaiabhayk]

The training hangs for JigsawPuzzleSolver, I used the following command to run :-

./caffe-master-jps/build/tools/caffe train --solver=solver_cfn_jps.prototxt --gpu=0,1

I0327 16:41:56.982856 15866 net.cpp:230] data_0 does not need backward computation.
I0327 16:41:56.982861 15866 net.cpp:272] This network produces output accuracy
I0327 16:41:56.982864 15866 net.cpp:272] This network produces output loss
I0327 16:41:57.006620 15866 net.cpp:285] Network initialization done.
I0327 16:41:57.007475 15866 solver.cpp:60] Solver scaffolding done.
I0327 16:41:57.025913 15866 parallel.cpp:391] GPUs pairs 0:1
I0327 16:41:57.375725 15866 data_transformer.cpp:50] hash size: 100 , 9
I0327 16:41:57.375762 15866 data_transformer.cpp:99] Patch Rectangles :
I0327 16:41:57.375768 15866 data_transformer.cpp:101] { 0 , 0 , 85 , 85 }
I0327 16:41:57.375773 15866 data_transformer.cpp:101] { 85 , 0 , 85 , 85 }
I0327 16:41:57.375778 15866 data_transformer.cpp:101] { 170 , 0 , 85 , 85 }
I0327 16:41:57.375782 15866 data_transformer.cpp:101] { 0 , 85 , 85 , 85 }
I0327 16:41:57.375787 15866 data_transformer.cpp:101] { 85 , 85 , 85 , 85 }
I0327 16:41:57.375790 15866 data_transformer.cpp:101] { 170 , 85 , 85 , 85 }
I0327 16:41:57.375794 15866 data_transformer.cpp:101] { 0 , 170 , 85 , 85 }
I0327 16:41:57.375799 15866 data_transformer.cpp:101] { 85 , 170 , 85 , 85 }
I0327 16:41:57.375803 15866 data_transformer.cpp:101] { 170 , 170 , 85 , 85 }

Any suggestion or is this a know issue ?