Hello, everyone. I’m currently training an image segmentation model using PyTorch.
While doing this, I’ve watched the training results like the below picture.
I thought the model would be better if I train it more. (Currently, I set the minimum number of epochs to 0 and the maximum number of epochs to 200, and used the early stopping technique.)
I wonder whether it is a good way to set the minimum number of epochs (like setting it to 30), and I also want to ask you about which value is proper for the minimum number of epochs.