Check failed: status == CUDNN_STATUS_SUCCESS (1 vs. 0) CUDNN_STATUS_NOT_INITIALIZED

NVIDIA / gpu-rest-engine

A REST API for Caffe using Docker and Go

BSD 3-Clause "New" or "Revised" License

421 stars 94 forks source link

Check failed: status == CUDNN_STATUS_SUCCESS (1 vs. 0) CUDNN_STATUS_NOT_INITIALIZED #23

Closed rperdon closed 6 years ago

rperdon commented 6 years ago

Installed CUDA 9.0 package and driver. Installed cudnn 7.0.

Checked install, of libcudnn, seems ok. LD_LIBRARY_PATH is set correctly I think and files seems to be ok spot. root@Hackathon:/home/cpcmecadmin/caffe/build# check libcudnn libcudnn.so.7 -> libcudnn.so.7.0.5 libcudnn.so.7 -> libcudnn.so.7.0.5 libcudnn.so.5 -> libcudnn.so.5.1.10 libcudnn is installed

When trying to run docker run --runtime=nvidia --name=server --net=host --rm inference_server I get the above error. Any thoughts?

flx42 commented 6 years ago

Oh, I messed up the Dockerfile I think, will fix.

flx42 commented 6 years ago

Fixed. Please test @rperdon