Thank you for sharing :) I tried your code with 700 plane images,
train loss can descend continuously, but validation loss can not descend after about 5 epochs.
Your readme notebook shows the same phenomenon, can you tell me how you overcome this problem ?
Thank you for sharing :) I tried your code with 700 plane images, train loss can descend continuously, but validation loss can not descend after about 5 epochs. Your readme notebook shows the same phenomenon, can you tell me how you overcome this problem ?