My implementation of the Flowers Images Classifier as the Final Lab Project of the PyTorch Challenge by Udacity.
With the VGG-19 net with Batch Normalization and 13 epochs processed (out of 40), the validation accuracy of 83% was reached. Due to inavailability of GPU, and, thus, very slow training (13 epochs in 24 hours), the model traning was stopped at the point of 83% accuracy on Validation Data.