FLUID: A Unified Evaluation Framework for Flexible Sequential Data

TL;DR

FLUID is a comprehensive evaluation framework that unifies multiple learning paradigms to assess and develop general machine learning methods capable of handling diverse, real-world data challenges in sequential settings.

Contribution

The paper introduces FLUID, a new unified evaluation framework that integrates few-shot, continual, transfer, and representation learning objectives for more realistic ML assessment.

Findings

New baselines outperform existing methods on FLUID.

Current solutions have specific advantages and limitations.

FLUID reveals new research challenges for general ML methods.

Abstract

Modern ML methods excel when training data is IID, large-scale, and well labeled. Learning in less ideal conditions remains an open challenge. The sub-fields of few-shot, continual, transfer, and representation learning have made substantial strides in learning under adverse conditions; each affording distinct advantages through methods and insights. These methods address different challenges such as data arriving sequentially or scarce training examples, however often the difficult conditions an ML system will face over its lifetime cannot be anticipated prior to deployment. Therefore, general ML systems which can handle the many challenges of learning in practical settings are needed. To foster research towards the goal of general ML methods, we introduce a new unified evaluation framework - FLUID (Flexible Sequential Data). FLUID integrates the objectives of few-shot, continual, transfer, and representation learning while enabling comparison and integration of techniques across these subfields. In FLUID, a learner faces a stream of data and must make sequential predictions while choosing how to update itself, adapt quickly to novel classes, and deal with changing data distributions; while accounting for the total amount of compute. We conduct experiments on a broad set of methods which shed new insight on the advantages and limitations of current solutions and indicate new research problems to solve. As a starting point towards more general methods, we present two new baselines which outperform other evaluated methods on FLUID. Project page: https://raivn.cs.washington.edu/projects/FLUID/.