Dataset parallelization for distributed nodes using Neural Network

I'm looking for a freelancer with experience in Python programming language, applying PyTorch/mpi4y or some other deep learning framework for dataset parallelization for distributed nodes. I will be providing the dataset and need expertise on the parallelization process, from distributed computing means parallelized datast program should run on multiple nodes.The successful freelancer should have extensive experience with Python, distributed programming such as mpi/cuda,PyTorch, and large datasets. The end result should be an automated process of mapping input data with Neural Network to achieve efficient parallelization.