no code implementations • 4 Jun 2019 • Andre Viebke, Sabri Pllana, Suejb Memeti, Joanna Kolodziej
We evaluate the prediction accuracy of performance models in the context of training three different convolutional neural network architectures on the Intel Xeon Phi.
no code implementations • 25 Feb 2017 • Andre Viebke, Suejb Memeti, Sabri Pllana, Ajith Abraham
In this paper, we present our parallelization scheme for training convolutional neural networks (CNN) named Controlled Hogwild with Arbitrary Order of Synchronization (CHAOS).