Search Results for author: Andrew Or

Found 1 papers, 0 papers with code

VirtualFlow: Decoupling Deep Learning Models from the Underlying Hardware

no code implementations20 Sep 2020 Andrew Or, Haoyu Zhang, Michael J. Freedman

In our evaluation, our implementation of VirtualFlow for TensorFlow achieved strong convergence guarantees across different hardware with out-of-the-box hyperparameters, up to 48% lower job completion times with resource elasticity, and up to 42% higher throughput with heterogeneous training.

Cannot find the paper you are looking for? You can Submit a new open access paper.