1: \begin{abstract}

2: Decentralized stochastic optimization methods have gained a lot of attention recently, mainly because of their cheap per iteration cost, data locality, and their communication-efficiency.

3: In this paper we introduce a unified convergence analysis that covers a large variety of decentralized SGD methods which so far have required different intuitions, have different applications, and which have been developed separately in various communities. 

4: \\

5: Our algorithmic framework covers

6: local SGD updates and synchronous and pairwise gossip updates on adaptive network topology.

7: We derive universal convergence rates for smooth (convex and non-convex) problems and the rates interpolate between the heterogeneous (non-identically distributed data) and iid-data settings, recovering linear convergence rates in many special cases, for instance for over-parametrized models. %

8: Our proofs rely on weak assumptions (typically improving over prior work in several aspects) and recover (and improve) the best known complexity results for a host of important scenarios, such as for instance

9: coorperative SGD and 

10: federated averaging (local SGD).

11: \end{abstract}

12: