search

pluskid / Mocha.jl

Deep Learning framework for Julia
Other
1.29k stars 254 forks source link

unable to init gpu back end. #167

Closed ashleylid closed 8 years ago

ashleylid commented 8 years ago

Kernels build fine. Setting the device works fine: ENV["MOCHA_USE_CUDA"] = "true" but backend = GPUBackend() throws this error: (is there something I am missing in the install?)

ERROR: LoadError: ccall: could not find function cudnnCreate in library [inlined code] from /home/ashley/.julia/v0.4/Mocha/src/cuda/cudnn.jl:51 in create at /home/ashley/.julia/v0.4/Mocha/src/cuda/cudnn.jl:62 in init at /home/ashley/.julia/v0.4/Mocha/src/cuda/backend.jl:166 in include at ./boot.jl:261 in include_from_node1 at ./loading.jl:304 in process_options at ./client.jl:280 in _start at ./client.jl:378 while loading /home/ashley/.julia/v0.4/Mocha/test/runtests.jl, in expression starting on line 18 ===============================================================================================================[ ERROR: Mocha ]================================================================================================================

failed process: Process(/usr/bin/julia --check-bounds=yes --code-coverage=none --color=yes /home/ashley/.julia/v0.4/Mocha/test/runtests.jl, ProcessExited(1)) [1]

ERROR: Mocha had test errors in error at ./error.jl:21 in test at pkg/entry.jl:803 in anonymous at pkg/dir.jl:31 in cd at file.jl:22 in cd at pkg/dir.jl:31 in test at pkg.jl:71

ashleylid commented 8 years ago

Thanks to issue #164 I was able to find my issue. I had to change the const link to libcudnn in the cudnn.jl script. Its now a hard link specific to my machine - but at least it works.