ByteTransformer: A High-Performance Transformer Boosted for Variable-Length Inputs-Reference-Cited by-同舟云学术

ByteTransformer: A High-Performance Transformer Boosted for Variable-Length Inputs

Published:2023-05 Issue: Volume: Page:
ISSN:
Container-title:2023 IEEE International Parallel and Distributed Processing Symposium (IPDPS)
language:
Short-container-title:

Author:

Zhai Yujia¹,Jiang Chengquan²,Wang Leyuan²,Jia Xiaoying²,Zhang Shang³,Chen Zizhong¹,Liu Xin²,Zhu Yibo²

Affiliation:

1. University of California,Riverside

2. ByteDance Ltd.

3. NVIDIA Corporation

Publisher

IEEE

Link

Reference37 articles.

2. Albert: A lite bert for self-supervised learning of language representations;lan,2019

Cited by 12 articles. 订阅此论文施引文献订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献

1. Arlo: Serving Transformer-based Language Models with Dynamic Input Lengths;Proceedings of the 53rd International Conference on Parallel Processing;2024-08-12

3. Optimizing Attention by Exploiting Data Reuse on ARM Multi-core CPUs;Proceedings of the 38th ACM International Conference on Supercomputing;2024-05-30

5. Optimizing Dynamic-Shape Neural Networks on Accelerators via On-the-Fly Micro-Kernel Polymerization;Proceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2;2024-04-27