# Different results between training and evaluation phase on the same data

I have trained a CNN and in the training phase I obtained an accuracy of 36.5%. If I call model.predict() on the same test data of the training phase I only obtain a 14.5% accuracy score. How is it possible ?

It's an 8 class classification problem.

• If it is in fact the same data with the same transforms, data augmentations, etc., I'd say it's not possible. But there is not enough information in your question. Please edit your post to add details and a reproducible example. Sep 2 at 10:43
• Esp. how accuracy is measured ... Sep 8 at 15:27