Distributed Neural Net Training
Training logistic regression model with multiple machines in parallel using MPI and CUDA. Achieved >5x speedup on training datasets, such as weather prediction.
Final Project for 15-418 Parallel Computer Architecture and Programming