Extreme Model Compression for On-device Natural Language Understanding

Kanthashree Mysore Sathyendra, Samridhi Choudhary, Leah Nicolich-Henkin


Abstract
In this paper, we propose and experiment with techniques for extreme compression of neural natural language understanding (NLU) models, making them suitable for execution on resource-constrained devices. We propose a task-aware, end-to-end compression approach that performs word-embedding compression jointly with NLU task learning. We show our results on a large-scale, commercial NLU system trained on a varied set of intents with huge vocabulary sizes. Our approach outperforms a range of baselines and achieves a compression rate of 97.4% with less than 3.7% degradation in predictive performance. Our analysis indicates that the signal from the downstream task is important for effective compression with minimal degradation in performance.
Anthology ID:
2020.coling-industry.15
Volume:
Proceedings of the 28th International Conference on Computational Linguistics: Industry Track
Month:
December
Year:
2020
Address:
Online
Editors:
Ann Clifton, Courtney Napoles
Venue:
COLING
SIG:
Publisher:
International Committee on Computational Linguistics
Note:
Pages:
160–171
Language:
URL:
https://aclanthology.org/2020.coling-industry.15
DOI:
10.18653/v1/2020.coling-industry.15
Bibkey:
Cite (ACL):
Kanthashree Mysore Sathyendra, Samridhi Choudhary, and Leah Nicolich-Henkin. 2020. Extreme Model Compression for On-device Natural Language Understanding. In Proceedings of the 28th International Conference on Computational Linguistics: Industry Track, pages 160–171, Online. International Committee on Computational Linguistics.
Cite (Informal):
Extreme Model Compression for On-device Natural Language Understanding (Mysore Sathyendra et al., COLING 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.coling-industry.15.pdf