Your question is Gradient Descent and Local Minima. 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're training a supervised learning model and need to explain how the optimizer updates parameters during training. You also want to reason about what can go wrong when the loss surface is non-convex.
Explain how gradient descent optimization works and how you would prevent a model from getting stuck in local minima.