Performance Trade-off Between Intermediate-Task Training and Direct Fine-Tuning on XGLUE for Deployment Efficiency

Supplementary Training on Intermediate Labeled-data Tasks (STILTs) is a widely applied technique, which first fine-tunes the pretrained language models on an intermediate task before on the target task of interest. While STILTs is able to further improve the performance of pretrained language models, it is still unclear why and when it works. Previous research shows that those intermediate tasks involving complex inference, such as commonsense reasoning, work especially well for RoBERTa. In this paper, we discover that the improvement from an intermediate task could be orthogonal to it contain

Research goal: What is the performance trade-off between intermediate-task training and direct target-task fine-tuning in terms of inference latency and accuracy on XGLUE, when optimizing for deployment efficiency?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.5/10.