Work

Consistent Accelerated Inference via Confident Adaptive Transformers

Year: 2021

Type: conference-paper

Abstract: We develop a novel approach for confidently accelerating inference in the large and expensive multilayer Transformers that are now ubiquitous in natural language processing (NLP).Amortized or approximate computational methods increase efficiency, but can come with unpredictable performance costs.In this work, we present CATs-Confident Adaptive Transformers-in which we simultaneously increase computational efficiency, while guaranteeing a specifiable degree of consistency with the original model with high confidence.Our method trains additional prediction heads on top of intermediate layers, and dynamically decides when to stop allocating computational effort to each input using a meta consistency classifier.To calibrate our early prediction stopping rule, we formulate a unique extension of conformal prediction.We demonstrate the effectiveness of this approach on four classification and regression tasks. 1 * The first two authors contributed equally. 1

Source: Conference on Empirical Methods in Natural Language Processing (EMNLP)

DOI: 10.18653/v1/2021.emnlp-main.406

Authors: Tal Schuster, Adam Fisch, Tommi Jaakkola, Regina Barzilay

Institutions: MIT Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology

Language: English

Topic: Topic Modeling

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funders: National Science Foundation, Department of Science and Technology, Ministry of Science and Technology, India, Defence Science and Technology Agency - Singapore

Award: G712593339