ACL 2023short3 citations
Self-Distilled Quantization: Achieving High Compression Rates in Transformer-Based Language Models
Abstract
We investigate the effects of post-training quantization and quantization-aware training on the generalization of Transformer language models. We present a new method called self-distilled quantization (SDQ) that minimizes accumulative quantization errors and outperforms baselines. We apply SDQ to multilingual models XLM-RBase and InfoXLMBase and demonstrate that both models can be reduced from 32-bit floating point weights to 8-bit integer weights while maintaining a high level of performance on the XGLUE benchmark. Our results also highlight the challenges of quantizing multilingual models, which must generalize to languages they were not fine-tuned on.
BibTeX
@inproceedings{oneill-dutta-2023-self,
title = "Self-Distilled Quantization: Achieving High Compression Rates in Transformer-Based Language Models",
author = "O{'}Neill, James and
Dutta, Sourav",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.acl-short.114/",
doi = "10.18653/v1/2023.acl-short.114",
pages = "1329--1339"
}