Comipling the CPU-only feature extractor

TorontoDeepLearning / convnet

A GPU implementation of Convolutional Neural Nets in C++

BSD 2-Clause "Simplified" License

504 stars 229 forks source link

Comipling the CPU-only feature extractor #25

Open tjusxh opened 9 years ago

tjusxh commented 9 years ago

According the install instruction. Set the paths to the dependencies in convnet/cpu/Makefile I don't find the file in convnet about the version.Please help me.

avdmitry commented 9 years ago

@tjusxh,

convnet/cpu/Makefile

Now it placed in convnet/apps/cpu/Makefile.

Also there is bug in CPU feature extractor, I created PR, which fix it: https://github.com/TorontoDeepLearning/convnet/pull/26

Not sure that I understand your question: "file in convnet about the version".

manosbar commented 9 years ago

Hi, what tjusxh is trying to say is that he cannot find the necessary Makefile. Same problem here. Actually, there is no convnet/cpu or convent/apps/cpu 2015-06-11 12 37 38 2015-06-11 12 38 30

yhancik commented 9 years ago

I think there's only one Makefile now, with a USE_CUDA flag. EDIT: but obviously (?) it only compiles the GPU version. At least that's what I assume (from the segfault it's giving me). I had better luck compiling from older version of the sources, before the merge of the makefiles.

It would be nice if the install documentation was updated, though :)