Can I add new variable to custom loss function? (tensorflow keras)

Question

I usually use the keras model with custom loss function. I want add variable which is controlled by optimizer (e.g., Adam) to the loss function, but nothing clear yet.

<