ERROR: Check failed: error == cudaSuccess (30 vs. 0) unknown error

[Do It]

I get this error when I try to create a new model in DIGITS. I went through the intro tutorial to DIGITS involving the MNIST and had no trouble loading the data set or creating/running the sample model.

When I try run a new model using AlexNet with my own images, it fails within 2 seconds and this error pops up:

"ERROR: Check failed: error == cudaSuccess (30 vs. 0) unknown error
Attempting to upgrade input file specified using deprecated 'solver_type' field (enum)': /var/lib/digits/jobs/20170622-130536-c681/solver.prototxt Successfully upgraded file specified using deprecated 'solver_type' field (enum) to 'type' field (string). Note that future Caffe releases will only support 'type' field (string) for a solver's type. Check failed: error == cudaSuccess (30 vs. 0) unknown error"

Any ideas on what might fix this? I tried the sudo modprobe command someone suggested and it didn't work. Other posts I've seen that have this error message were in different contexts.

[Object Detection is My Game]

What are the specs of your server, e.g. GPU and it's compute capability?

Take a look at this forum: https://stackoverflow.com/questions/38542763/how-can-i-make-tensorflow-run-on-a-gpu-with-capability-2-0

And then search for your CUDA GPU version: https://developer.nvidia.com/cuda-gpus

cuDNN will not work with compute capabilities lower than 3.0

[naj...@axelspace.com]

I would say this is a output problem. Make sure that the 'num_output' (in .prototxt file) of the the last layer matches the number of classes (labels) of your dataset.