Hi I have a quick question about overfitting.
In the graph above, yellow is baseline, and what I changed after modifying the model is gray. If you look at it, my model(gray) is very good at first, but it catches up and gets bad results. Can we call this overfitting?
I don’t think the gray one is learning wrong. (perhaps not 
) How can I change the training schedule? Is there any other things to consider?
As per the first link that turned up when I asked Google about overfitting, overfitting “occurs when a statistical model fits exactly against its training data”. If you think that this is happening in your picture, then you can call it overfitting.