recipe for target 'obj/convolutional_kernels.o' failed

[Geobm]

Hi @nicholaskajoh,

I would like to use the demo of your repo but I'm gettin the following error when running make once I cloned the darknet repo and installed its dependencies in order to use YOLO. Also, I am using Ubuntu 18.04.1.

$ sudo make

nvcc  -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DGPU -I/usr/local/cuda/include/ --compiler-options "-Wall -Wno-unused-result -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DGPU" -c ./src/convolutional_kernels.cu -o obj/convolutional_kernels.o
nvcc fatal   : Path to libdevice library not specified
Makefile:92: recipe for target 'obj/convolutional_kernels.o' failed
make: *** [obj/convolutional_kernels.o] Error 1

I supposed it could be something wrong with CUDA's path as suggested by this comment. Since the path in the Makefile its the same in /usr/local/cuda/include/

ifeq ($(GPU), 1) 
COMMON+= -DGPU -I/usr/local/cuda/include/
CFLAGS+= -DGPU
LDFLAGS+= -L/usr/local/cuda/lib64 -lcuda -lcudart -lcublas -lcurand
endif

Tried the following command export PATH=/usr/local/cuda-8.0/bin${PATH:+:${PATH}} as mentioned but could not fix it.

Any suggestion of what I'm doing wrong will be appreciated

[nicholaskajoh]

You don't need to install darknet. OpenCV has a DNN module that can handle darknet models. Once you install the dependencies in requirements.txt, you're good.