On the Convergence of Adam under Non-uniform Smoothness: Separability from SGDM and Beyond
This work clarifies theoretical distinctions between Adam and SGDM for optimization researchers, though it is incremental as it builds on existing convergence analyses.
This paper demonstrates that Adam achieves faster convergence than Stochastic Gradient Descent with Momentum (SGDM) under non-uniform smoothness, matching lower bounds for deterministic and stochastic first-order optimizers, while SGDM can fail to converge in some cases.
This paper aims to clearly distinguish between Stochastic Gradient Descent with Momentum (SGDM) and Adam in terms of their convergence rates. We demonstrate that Adam achieves a faster convergence compared to SGDM under the condition of non-uniformly bounded smoothness. Our findings reveal that: (1) in deterministic environments, Adam can attain the known lower bound for the convergence rate of deterministic first-order optimizers, whereas the convergence rate of Gradient Descent with Momentum (GDM) has higher order dependence on the initial function value; (2) in stochastic setting, Adam's convergence rate upper bound matches the lower bounds of stochastic first-order optimizers, considering both the initial function value and the final error, whereas there are instances where SGDM fails to converge with any learning rate. These insights distinctly differentiate Adam and SGDM regarding their convergence rates. Additionally, by introducing a novel stopping-time based technique, we further prove that if we consider the minimum gradient norm during iterations, the corresponding convergence rate can match the lower bounds across all problem hyperparameters. The technique can also help proving that Adam with a specific hyperparameter scheduler is parameter-agnostic, which hence can be of independent interest.