Horn: A System for Parallel Training and Regularizing of Large-Scale Neural Networks
This addresses the challenge of scaling neural network training for researchers and practitioners, but it appears incremental as it builds on existing distributed training and regularization techniques.
The paper tackles the problem of training and regularizing large-scale neural networks on distributed systems by introducing a new system with flexible model partitioning and parallelization strategies, achieving effectiveness demonstrated on MNIST handwritten digits classification.
I introduce a new distributed system for effective training and regularizing of Large-Scale Neural Networks on distributed computing architectures. The experiments demonstrate the effectiveness of flexible model partitioning and parallelization strategies based on neuron-centric computation model, with an implementation of the collective and parallel dropout neural networks training. Experiments are performed on MNIST handwritten digits classification including results.