Distributed Optimization for Over-Parameterized Learning
arXiv:1906.06205
Abstract
Distributed optimization often consists of two updating phases: local optimization and inter-node communication. Conventional approaches require working nodes to communicate with the server every one or few iterations to guarantee convergence. In this paper, we establish a completely different conclusion that each node can perform an arbitrary number of local optimization steps before communication. Moreover, we show that the more local updating can reduce the overall communication, even for an infinity number of steps where each node is free to update its local model to near-optimality before exchanging information. The extra assumption we make is that the optimal sets of local loss functions have a non-empty intersection, which is inspired by the over-paramterization phenomenon in large-scale optimization and deep learning. Our theoretical findings are confirmed by both distributed convex optimization and deep learning experiments.
References in corpus (10)
- Understanding deep learning requires rethinking generalization
- QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding
- A Convergence Theory for Deep Learning via Over-Parameterization
- Gradient Descent Provably Optimizes Over-parameterized Neural Networks
- Local SGD Converges Fast and Communicates Little
- On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization
- The Power of Interpolation: Understanding the Effectiveness of SGD in Modern Over-parametrized Learning
- Parallel SGD: When does averaging help?
- On exponential convergence of SGD in non-convex over-parametrized learning
- An Optimal Control Approach to Deep Learning and Applications to Discrete-Weight Neural Networks