This repository provides a Keras implementation of SqueezeNet v1.1, a convolutional neural network architecture designed for efficient image classification. The implementation uses Keras Functional Framework 2.0 and includes pretrained weights converted from the original Caffe model. The model achieves AlexNet-level accuracy with a significantly smaller model size of 5.1 MB. Users can load the model and perform predictions on images using standard Keras preprocessing utilities.