Virtual neural networks: hundreds of souls in a body This repository presents relevant source code for the paper Virtual neural networks: hundreds of souls in a body. This paper introduces a novel paradigm called virtual neural networks, where the concept of an ensemble approach involving many 'virtual' models that share weights determined by a few 'physical' models. The ensemble consists of up to hundreds of virtual models that are trained concurrently. Moreover, all virtual networks share the same input, and their complex structure creates a form of inner augmentation that enhances the robustness of the entire ensemble. A more detailed explanation is available in the paper Virtual neural networks: hundreds of souls in a body (full reference to the article will be provided after the publication). The implementation converts virtual models for the EfficientNet architecture. The implementation is realized in Python 3.8 using IPython notebooks.