A research and development team is training a custom PyTorch model in Amazon SageMaker AI. The team wants to implement a solution where each member will be alerted through email notifications when the model is overfitting. Which solution should the team implement?