Context Aim here is to leverage pretrained network so that our algorithms work more efficiently and quick. Content It is a pretrained network called MobilNet. It accepts 128x128 images and examines them resulting in image classifications over 1000 classes. It has to be used without top layers because only 224x224 images can be used with top layers. So one should manipulate top layers depending on his own intention. Acknowledgements This pretrained network is available on "https://github.com/fchollet/deep-learning-models/releases" together with other pretrained networks. Thanks to user "fchollet" of github for publishing all of them in one place. Inspiration