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

Open in new window