I was training a Deep Neural Network using transfer learning and the loss value as of now is close to 1.7. I am using a dataset of 31,000 images and I do apply data augmentation techniques to help the model generalize better. However, I had confusion regarding epochs and loss values. Should I be training the model for say 50 epochs for good accuracy or would it be better if I trained until the loss value is as low as possible which will be very close to zero. Which one is recommended to get the best accuracy? Will more epochs cause the model to perform better than the lowest loss value? Any help would be greatly appreciated. Thanks in advance!
I think it is better to use early stopping method.Too many epochs will lead to overfitting the model.Early stopping allows you to stop the training once the model performance stops improving on a hold out validation dataset.
refer this link to the concept
Plot the loss of both train amd test loss on single graph after every epoch and check that it should not overfit or underfit..most commenly overfit happens in deep learning so keep track on that. Once the your distance between both train and test loss increases on graph stop at that time and save the weights. You can also keep track of weights of every layer also if your model overfit.