No OpKernel was registered to support Op 'FarthestPointSample' with these attrs. Registered devices: [CPU], Registered kernels: device='GPU'

[GopiRajuMatta]

Hello,

My installation was successful. I setup the environment using environment.yml file. Previously, I ran this code of testing and training without any errors. Suddenly, I came across this error, while running test code.

"InvalidArgumentError (see above for traceback): Restoring from checkpoint failed. This is most likely due to a mismatch between the current graph and the graph from the checkpoint. Please ensure that you have not altered the graph expected based on the checkpoint. Original error:

No OpKernel was registered to support Op 'FarthestPointSample' with these attrs. Registered devices: [CPU], Registered kernels:
device='GPU' "

Can you please help me out to resolve this error.

Thank you
Gopi

[GopiRajuMatta]

@yifita , can you respond to this.
When I searched in internet, people are saying like it is problem with GPU. While running, it is picking up CPU instead GPU. I uninstalled and installed all again, but same problem. Previously it worked without any problems. I don't know what is the problem particularly happening now. Looking forward for your reply.

If anyone came across same issue, please respond.

[Nhunguts]

I met the same problem. If I uninstall tensorflow and reinstall tensorflow-gpu, this error has gone, but another error occured:

ImportError: libcudnn.so.6: cannot open shared object file: No such file or directory
Failed to load the native TensorFlow runtime.

I am using tensorflow 1.3.0 and cuda8.0.

[GopiRajuMatta]

This post would be helpful!
Franck-Dernoncourt/NeuroNER#66