Model Accuracy: 0.9754 Validation Accuracy: 0.8285
Dataset: https://www.kaggle.com/tongpython/cat-and-dog
The model is already trained on the dataset and trained model is also included in this repo.
To retrain the model: Feel free to add/remove/change the layers of the network and retrain and save the model by calling the function model_training() or simply remove comment on line 62 of cnn.py.
To predict any custom image: Replace "file_destination_here" with file path of the image which you want to predict at line 76 of cnn.py.