How am I suppose to know if my deep learning model is ready to be tested based on evaluation metrics

Question

I am currently working on a skin cancer classification problem. The data consists of 9 target values and the model is trained with densenet for 20 epochs. The model file has