Conditional Adapters: Parameter-efficient Transfer Learning with Fast Inference

Tao Lei, Junwen Bai, Siddhartha Brahma, Joshua Ainslie, Kenton Lee, Yanqi Zhou, Nan Du, Vincent Y Zhao, Yuexin Wu, Bo Li, Yu Zhang, Ming-Wei Chang

Keywords: conditional computation, inference efficiency, parameter efficiency, large models

TL;DR: a parameter-efficient transfer learning method that also improves inference efficiency

Abstract:

We propose Conditional Adapter (CoDA), a parameter-efficient transfer learning method that also improves inference efficiency. CoDA generalizes beyond standard adapter approaches to enable a new way of balancing speed and accuracy using conditional computation.
Starting with an existing dense pretrained model, CoDA adds sparse activation together with a small number of new parameters and a light-weight training phase.
Our experiments demonstrate that the CoDA approach provides an unexpectedly efficient way to transfer knowledge.
Across a variety of language, vision, and speech tasks, CoDA achieves a 2x to 8x inference speed-up compared to the state-of-the-art Adapter approaches with moderate to no accuracy loss and the same parameter efficiency.

Submission Number: 14632