Matplotlib works for non NN tasks but kernel dies when used to plot accuracy and loss from tf keras NN. Using mac and re-installed anaconda, tensorflow 2.0, keras 2.3.1, matplotlib 3.1.3, numpy 1.18.1 and pandas 1.0.3. Any help would be appreciated. Thank you.
Sorry but I think you’re going to have to ask Keras for help as it sounds like a bug on their end. Can you post some code for reproduction?
Hey @Hongmeng ,
Kernel usally dies when you are RAM overflows or when there is an interrupt present in your code. Matplotlib however as far as I know does not introduce any of these hurdles so i would suggest you to go through your model architecture one more time and see if you find some errors in it.