search

jcjohnson / neural-style

Torch implementation of neural style algorithm
MIT License
18.31k stars 2.7k forks source link

THCudaCheck FAIL line=62 error=35 #355

Open krwarner opened 7 years ago

krwarner commented 7 years ago

When I input sudo th neural_style.lua -style_image sao1.jpg -content_image nge_02.jpg -gpu -1

I get the output THCudaCheck FAIL file=/tmp/luarocks_cutorch-scm-1-8295/cutorch/lib/THC/THCGeneral.c line=62 error=35 : CUDA driver version is insufficient for CUDA runtime version /usr/local/bin/luajit: /usr/local/share/lua/5.1/trepl/init.lua:389: cuda runtime error (35) : CUDA driver version is insufficient for CUDA runtime version at /tmp/luarocks_cutorch-scm-1-8295/cutorch/lib/THC/THCGeneral.c:62 stack traceback: [C]: in function 'error' /usr/local/share/lua/5.1/trepl/init.lua:389: in function 'require' neural_style.lua:329: in function 'setup_gpu' neural_style.lua:53: in function 'main' neural_style.lua:604: in main chunk [C]: in function 'dofile' /usr/local/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk [C]: at 0x00406260 Thank you for the help!

MrZoidberg commented 7 years ago

CUDA driver version is insufficient for CUDA runtime version. Check your CUDA driver and runtime version. This guide might be helpful.

krwarner commented 7 years ago

I don't have a CUDA compatible video card. However, that shouldn't matter as I input -gpu -1 in order to run on only CPU and not require CUDA or GPU.

jcjohnson commented 7 years ago

@krwarner Can you try updating neural-style? I think this error is due to a bug that was fixed in https://github.com/jcjohnson/neural-style/commit/ebccf67b283e5ca8f2ef6799332c63740556937a.

Annachen0918 commented 7 years ago

@jcjohnson hey johnson, i have updated neural-style and kept getting the output exactly same with @krwarner

kotodharma commented 6 years ago

Bless you, krwarner, the -gpu flag was exactly what I needed to get beyond my CUDA errors!