Issue with running torch (cuda error)

I have this issue that I don’t know what’s the fix here.

/home/phong/torch/install/bin/lua: /home/phong/torch/install/share/lua/5.2/nn/Linear.lua:66: cublas runtime error : an internal operation failed at /home/phong/torch/extra/cutorch/lib/THC/THCBlas.cu:246
stack traceback:
[C]: in function ‘addmm’
/home/phong/torch/install/share/lua/5.2/nn/Linear.lua:66: in function </home/phong/torch/install/share/lua/5.2/nn/Linear.lua:53>
(…tail calls…)
/home/phong/torch/install/share/lua/5.2/nngraph/gmodule.lua:345: in function ‘neteval’
/home/phong/torch/install/share/lua/5.2/nngraph/gmodule.lua:380: in function </home/phong/torch/install/share/lua/5.2/nngraph/gmodule.lua:300>
(…tail calls…)
./onmt/modules/Network.lua:11: in function <./onmt/modules/Network.lua:10>
(…tail calls…)
/home/phong/torch/install/share/lua/5.2/nngraph/gmodule.lua:345: in function ‘neteval’
/home/phong/torch/install/share/lua/5.2/nngraph/gmodule.lua:380: in function </home/phong/torch/install/share/lua/5.2/nngraph/gmodule.lua:300>
(…tail calls…)
./onmt/modules/Decoder.lua:257: in function ‘forwardOne’
./onmt/modules/Decoder.lua:295: in function ‘forwardAndApply’
./onmt/modules/Decoder.lua:321: in function ‘forward’
./onmt/Seq2Seq.lua:178: in function ‘trainNetwork’
./onmt/train/Trainer.lua:193: in function ‘closure’
./onmt/utils/Parallel.lua:79: in function ‘launch’
./onmt/train/Trainer.lua:180: in function ‘trainEpoch’
./onmt/train/Trainer.lua:334: in function ‘train’
train.lua:150: in function ‘main’
train.lua:155: in main chunk
[C]: in function ‘dofile’
…hong/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
[C]: in ?