It gives an error over and over again although I restated the kernel as follows.
Did you change the parameters of the training? like batch size? or the optimization step?
I tried after an hour. It worked well. Thank you for your consideration 
1 Like