Too Brittle to Touch: Comparing the Stability of Quantization and Distillation towards Developing Low-Resource MT Models

Harshita Diddee, Sandipan Dandapat, Monojit Choudhury, Tanuja Ganu, Kalika Bali


Abstract
Leveraging shared learning through Massively Multilingual Models, state-of-the-art Machine translation (MT) models are often able to adapt to the paucity of data for low-resource languages. However, this performance comes at the cost of significantly bloated models which aren’t practically deployable. Knowledge Distillation is one popular technique to develop competitive lightweight models: In this work, we first evaluate its use in compressing MT models, focusing specifically on languages with extremely limited training data. Through our analysis across 8 languages, we find that the variance in the performance of the distilled models due to their dependence on priors including the amount of synthetic data used for distillation, the student architecture, training hyper-parameters and confidence of the teacher models, makes distillation a brittle compression mechanism. To mitigate this, we further explore the use of post-training quantization for the compression of these models. Here, we find that while Distillation provides gains across some low-resource languages, Quantization provides more consistent performance trends for the entire range of languages, especially the lowest-resource languages in our target set.
Anthology ID:
2022.wmt-1.80
Volume:
Proceedings of the Seventh Conference on Machine Translation (WMT)
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates (Hybrid)
Editors:
Philipp Koehn, Loïc Barrault, Ondřej Bojar, Fethi Bougares, Rajen Chatterjee, Marta R. Costa-jussà, Christian Federmann, Mark Fishel, Alexander Fraser, Markus Freitag, Yvette Graham, Roman Grundkiewicz, Paco Guzman, Barry Haddow, Matthias Huck, Antonio Jimeno Yepes, Tom Kocmi, André Martins, Makoto Morishita, Christof Monz, Masaaki Nagata, Toshiaki Nakazawa, Matteo Negri, Aurélie Névéol, Mariana Neves, Martin Popel, Marco Turchi, Marcos Zampieri
Venue:
WMT
SIG:
SIGMT
Publisher:
Association for Computational Linguistics
Note:
Pages:
870–885
Language:
URL:
https://aclanthology.org/2022.wmt-1.80
DOI:
Bibkey:
Cite (ACL):
Harshita Diddee, Sandipan Dandapat, Monojit Choudhury, Tanuja Ganu, and Kalika Bali. 2022. Too Brittle to Touch: Comparing the Stability of Quantization and Distillation towards Developing Low-Resource MT Models. In Proceedings of the Seventh Conference on Machine Translation (WMT), pages 870–885, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics.
Cite (Informal):
Too Brittle to Touch: Comparing the Stability of Quantization and Distillation towards Developing Low-Resource MT Models (Diddee et al., WMT 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.wmt-1.80.pdf