Your question is Measure Model Overfitting During Training. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are training a model and want to know whether it is starting to memorize the training data instead of learning patterns that generalize. The team wants a simple way to tell if performance gains during training are real or just overfitting.
How would you measure whether a model is overfitting during training?