I am getting the strange results from segmentation models.
I am using the same data as train and validation, the loss ( accuracy metric ) are not following each other, in raining it is going down (up) and for validation it is going up (down), for loss (accuracy).
Is there any difference between training and validation loss calculations?
Hi,
I am getting the strange results from segmentation models. I am using the same data as train and validation, the loss ( accuracy metric ) are not following each other, in raining it is going down (up) and for validation it is going up (down), for loss (accuracy). Is there any difference between training and validation loss calculations?
Thanks in advance!