Flag Aggregator: Scalable Distributed Training under Failures and Augmented Losses using Convex Optimization.

Hamidreza Almasi, Harsh Mishra, Balajee Vamanan, Sathya N. Ravi

Browse the full ICLR paper archive.