Model parameter management is essential for distributed training of large machine learning (ML) tasks. Some ML tasks are hard to distribute because common approaches to parameter management can be highly inefficient. Advanced parameter management approaches---such as selective replication or dynamic parameter allocation---can improve efficiency, but they typically need to be integrated manually into each task's implementation and they require expensive upfront experimentation to tune correctly. In this work, we explore whether these two problems can be avoided. We first propose a novel intent signaling mechanism that integrates naturally into existing ML stacks and provides the parameter manager with crucial information about parameter accesses. We then describe AdaPM, a fully adaptive, zero-tuning parameter manager based on this mechanism. In contrast to prior parameter managers, our approach decouples how access information is provided (simple) from how and when it is exploited (hard). In our experimental evaluation, AdaPM matched or outperformed state-of-the-art parameter managers out of the box, suggesting that automatic parameter management is possible.

Parameter servers (PSs) facilitate the implementation of distributed training for large machine learning tasks. In this paper, we argue that existing PSs are inefficient for tasks that exhibit non-uniform parameter access; their performance may even fall behind that of single node baselines. We identify two major sources of such non-uniform access: skew and sampling. Existing PSs are ill-suited for managing skew because they uniformly apply the same parameter management technique to all parameters. They are inefficient for sampling because the PS is oblivious to the associated randomized accesses and cannot exploit locality. To overcome these performance limitations, we introduce NuPS, a novel PS architecture that (i) integrates multiple management techniques and employs a suitable technique for each parameter and (ii) supports sampling directly via suitable sampling primitives and sampling schemes that allow for a controlled quality-efficiency trade-off. In our experimental study, NuPS outperformed existing PSs by up to one order of magnitude and provided up to linear scalability across multiple machine learning tasks.

Parameter servers (PSs) ease the implementation of distributed machine learning systems, but their performance can fall behind that of single machine baselines due to communication overhead. We demonstrate Lapse, an open source PS with dynamic parameter allocation . Previous work has shown that dynamic parameter allocation can improve PS performance by up to two orders of magnitude and lead to near-linear speed-ups over single machine baselines. This demonstration illustrates how Lapse is used and why it can provide order-of-magnitude speed-ups over other PSs. To do so, this demonstration interactively analyzes and visualizes how dynamic parameter allocation looks like in action.

We study scalable algorithms for frequent sequence mining under flexible subsequence constraints. Such constraints enable applications to specify concisely which patterns are of interest and which are not. We focus on the bulk synchronous parallel model with one round of communication; this model is suitable for platforms such as MapReduce or Spark. We derive a general framework for frequent sequence mining under this model and propose the D-SEQ and D-CAND algorithms within this framework. The algorithms differ in what data are communicated and how computation is split up among workers. To the best of our knowledge, D-SEQ and D-CAND are the first scalable algorithms for frequent sequence mining with flexible constraints. We conducted an experimental study on multiple real-world datasets that suggests that our algorithms scale nearly linearly, outperform common baselines, and offer acceptable generalization overhead over existing, less general mining algorithms.