Need help in Landslide Detection using deep learning and CNN

I have created a project on “Landslide detection using deep learning and CNN” but the model is overfitting. The validation accuracy is less than the training accuracy. Due to which the prediction is wrong. the model is classifying the image of dog or cat as landslide.

The basic idea of project is to determine landslide or non-landslide by feeding optical images to the machine. I’ve taken the training dataset of about 2000 images and test dataset of around 9 images.
but i’m unable to remove the overfitting.

please help!!

this is the link of my colab notebook:
https://colab.research.google.com/drive/176q_8MmuKEtDJ6SZ3WasfQ-aJbOvCyVv