CrossTrainer: Practical Domain Adaptation with Loss Reweighting

TL;DR

CrossTrainer is a practical domain adaptation system that uses loss reweighting to achieve high accuracy across datasets, with optimizations that reduce hyperparameter tuning costs and training time.

Contribution

The paper introduces CrossTrainer, a system that applies loss reweighting for domain adaptation and offers optimization techniques to improve efficiency.

Findings

High model accuracy across diverse datasets

Loss reweighting sensitivity to hyperparameters

Optimizations reduce training time and tuning effort

Abstract

Domain adaptation provides a powerful set of model training techniques given domain-specific training data and supplemental data with unknown relevance. The techniques are useful when users need to develop models with data from varying sources, of varying quality, or from different time ranges. We build CrossTrainer, a system for practical domain adaptation. CrossTrainer utilizes loss reweighting, which provides consistently high model accuracy across a variety of datasets in our empirical analysis. However, loss reweighting is sensitive to the choice of a weight hyperparameter that is expensive to tune. We develop optimizations leveraging unique properties of loss reweighting that allow CrossTrainer to output accurate models while improving training time compared to naive hyperparameter search.