Layer-Condensed KV Cache for Efficient Inference of Large Language Models

Haoyi Wu, Kewei Tu


Abstract
Huge memory consumption has been a major bottleneck for deploying high-throughput large language models in real-world applications. In addition to the large number of parameters, the key-value (KV) cache for the attention mechanism in the transformer architecture consumes a significant amount of memory, especially when the number of layers is large for deep language models. In this paper, we propose a novel method that only computes and caches the KVs of a small number of layers, thus significantly saving memory consumption and improving inference throughput. Our experiments on large language models show that our method achieves up to 26× higher throughput than standard transformers and competitive performance in language modeling and downstream tasks. In addition, our method is orthogonal to existing transformer memory-saving techniques, so it is straightforward to integrate them with our model, achieving further improvement in inference efficiency. Our code is available at https://github.com/whyNLP/LCKV.
Anthology ID:
2024.acl-long.602
Volume:
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
11175–11188
Language:
URL:
https://aclanthology.org/2024.acl-long.602
DOI:
10.18653/v1/2024.acl-long.602
Bibkey:
Cite (ACL):
Haoyi Wu and Kewei Tu. 2024. Layer-Condensed KV Cache for Efficient Inference of Large Language Models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 11175–11188, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Layer-Condensed KV Cache for Efficient Inference of Large Language Models (Wu & Tu, ACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.acl-long.602.pdf