Evaluating Your Model Exercise

Now that you've learned how to build CNN architectures and fine-tune them, let's learn more about how to evaluate their performance over training epochs. This is important in helping us know if our model is learning, and when to stop training.

We'll continue building off of the previous exercises, with the image augmentation code and one model architecture already written for you in the notebook below. Your job in this exercise is to build a function that will plot the training history to look at how:

• Training loss
• Validation loss
• Training accuracy
• Validation accuracy
  Change over the course of your training epochs.

Once you build this function, change the parameters of your model (learning rate and amount of dropout) to see how your model performance curves differ with different parameters.