I am training a model and my training accuracy is increasing almost linearly however the validation accuracy has constant accuracy with each epoch . I will do data augmentation to check whether the overfitting is resolved . Do I need to do data augmentation on the "Validation set " as well?
The reason for augmenting training data is for the model to better learn from a wider distribution of training data.
Don’t shuffle or augment validation data.
Please fix your model.