In this project, we try to reproduce the main findings based on paper EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.
The original paper's experiments were mainly conducted using the ImageNet dataset. In our project, we sought to reproduce the experiments using the CIFAR-10 and CIFAR-100 datasets. Please download the file function.py and model.py from main/models and import them while using efficientnet notebook.