Yathartha19/Image-Classification-Using-Deep-Learning — reverse-engineered prompt

Build me a simple Python deep learning project that can tell whether a photo is a cat or a dog. I want it to train a CNN inspired by VGG, using small stacked convolution layers, dropout, RMSprop, binary cross entropy, and basic image augmentation like rotation, zoom, shear, and horizontal flipping.

Use the Kaggle Dogs vs Cats dataset with an even split of cat and dog images, resize everything to 128 by 128, train the model, show validation accuracy, and save anything needed so I can reuse the trained model later. Please include a way to label or organize the images, train and test the model, plot the training results, and run a prediction on a new image so I can see if it says cat or dog.

Keep the code beginner friendly and organized into separate Python files. Add clear setup instructions and requirements so I can run it locally. Look up current docs online if you need to.