It is a unique way to train DL models using decentralised computing & Zero-Knowledge proofs for enhanced security & faster computations based on trustlessness & ultra-privacy.

Federated Learning is a privacy-preserving scheme to train deep learning models. Data exists in isolated pools and clients that are part of the network train a model with base parameters on their data. They share the updated model parameters with an aggregator that takes the federated average of this set of models. The result is going to be a new updated base model for the next epoch of training.

To remove the dependency on the server, we leverage ZK-Proofs to make the server trustless. The Zk-Proofs are then shown publicly so that anyone can verify whether or not the computation was done correctly.