Andrew Lavin

Hi @kashif I was implementing cuDNN v4 in this fork: https://github.com/andravin/cudnn-python-wrappers Some things are probably not implemented, been a couple months since I looked at it.

The first thing to check for each platform is whether the library name specified here is correct: ``` if sys.platform in ('linux2', 'linux'): _libcudnn_libname_list = ['libcudnn.so', 'libcudnn.so.4', 'libcudnn.so.4.0.4'] elif sys.platform...

I am not actively using it either, but I am happy to help out when I can. I want to stress that I really only tested the forward path convolution...

For reference, see https://github.com/terrychenism/OctaveConv/issues/4

OK cool, mostly wanted a sanity check in case I had missed something. I will take a crack at fixing it myself. Thanks for the quick feedback.

I am still using CUDA 6.5 and Driver version 343.19. Interesting that only the cuDNN tests fail for you. What graphics card are you using? I will upgrade to CUDA...

Hm, works for me with CUDA 7.0, driver 346.46. andrew@clive:~/develop/maxDNN/maxdnn$ ldd maxdnn_test.bin linux-vdso.so.1 => (0x00007fffc83dd000) libcuda.so.1 => /usr/lib/x86_64-linux-gnu/libcuda.so.1 (0x00007fa04fda7000) libcudart.so.7.0 => /usr/local/cuda/lib64/libcudart.so.7.0 (0x00007fa04fb49000) libnvToolsExt.so.1 => /usr/local/cuda/lib64/libnvToolsExt.so.1 (0x00007fa04f940000) libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2...

The unreasonable amount of memory in this case being size_t(-1) :-) Let's follow Scott's advice and try CUDNN_CONVOLUTION_FWD_NO_WORKSPACE by commenting out line convolution_cudnn.cpp:90 and uncommenting the next line.

My gtx980 is 4gb, but we have different maxwell chips. Yours is gm107 I believe while mine is gm204. Thanks for giving me test results for this gpu. So it...

I created a fix in my development branch and issued the above pull request. Can somebody with a GM107 verify? By the way, the insane workspace size is not exactly...