Asynchronous Stochastic Variational Inference
This work addresses the need for faster Bayesian computation in large-scale data settings, representing an incremental improvement by parallelizing an existing method.
The paper tackles the problem of scaling stochastic variational inference (SVI) to massive data by proposing a lock-free parallel implementation that allows asynchronous distributed computations, achieving linear speed-up with an asymptotic convergence rate of O(1/√T) while maintaining comparable performance to serial SVI.
Stochastic variational inference (SVI) employs stochastic optimization to scale up Bayesian computation to massive data. Since SVI is at its core a stochastic gradient-based algorithm, horizontal parallelism can be harnessed to allow larger scale inference. We propose a lock-free parallel implementation for SVI which allows distributed computations over multiple slaves in an asynchronous style. We show that our implementation leads to linear speed-up while guaranteeing an asymptotic ergodic convergence rate $O(1/\sqrt(T)$ ) given that the number of slaves is bounded by $\sqrt(T)$ ($T$ is the total number of iterations). The implementation is done in a high-performance computing (HPC) environment using message passing interface (MPI) for python (MPI4py). The extensive empirical evaluation shows that our parallel SVI is lossless, performing comparably well to its counterpart serial SVI with linear speed-up.