Exceeding the Limits of Visual-Linguistic Multi-Task Learning

TL;DR

This paper demonstrates that large-scale multi-task learning with over 1000 tasks across e-commerce product data can surpass previous limits, using a multi-modal transformer and novel heuristics for task-specific capacity allocation.

Contribution

It introduces a scalable methodology for multi-task learning on thousands of tasks, including new heuristics like DyPa for efficient parameter allocation.

Findings

Successful training of a single model on 1000 tasks

Identification of best practices for large-scale MTL

Introduction of DyPa heuristic for task-specific capacity

Abstract

By leveraging large amounts of product data collected across hundreds of live e-commerce websites, we construct 1000 unique classification tasks that share similarly-structured input data, comprised of both text and images. These classification tasks focus on learning the product hierarchy of different e-commerce websites, causing many of them to be correlated. Adopting a multi-modal transformer model, we solve these tasks in unison using multi-task learning (MTL). Extensive experiments are presented over an initial 100-task dataset to reveal best practices for "large-scale MTL" (i.e., MTL with more than 100 tasks). From these experiments, a final, unified methodology is derived, which is composed of both best practices and new proposals such as DyPa, a simple heuristic for automatically allocating task-specific parameters to tasks that could benefit from extra capacity. Using our large-scale MTL methodology, we successfully train a single model across all 1000 tasks in our dataset while using minimal task specific parameters, thereby showing that it is possible to extend several orders of magnitude beyond current efforts in MTL.