English

ECHO-LLaMA: Efficient Caching for High-Performance LLaMA Training

Machine Learning 2025-06-24 v2 Computation and Language

Abstract

This paper introduces ECHO-LLaMA, an efficient LLaMA architecture designed to improve both the training speed and inference throughput of LLaMA architectures while maintaining its learning capacity. ECHO-LLaMA transforms LLaMA models into shared KV caching across certain layers, significantly reducing KV computational complexity while maintaining or improving language performance. Experimental results demonstrate that ECHO-LLaMA achieves up to 77\% higher token-per-second throughput during training, up to 16\% higher Model FLOPs Utilization (MFU), and up to 14\% lower loss when trained on an equal number of tokens. Furthermore, on the 1.1B model, ECHO-LLaMA delivers approximately 7\% higher test-time throughput compared to the baseline. By introducing a computationally efficient adaptation mechanism, ECHO-LLaMA offers a scalable and cost-effective solution for pretraining and finetuning large language models, enabling faster and more resource-efficient training without compromising performance.

Keywords

Cite

@article{arxiv.2505.17331,
  title  = {ECHO-LLaMA: Efficient Caching for High-Performance LLaMA Training},
  author = {Maryam Dialameh and Rezaul Karim and Hossein Rajabzadeh and Omar Mohamed Awad and Hyock Ju Kwon and Boxing Chen and Walid Ahmed and Yang Liu},
  journal= {arXiv preprint arXiv:2505.17331},
  year   = {2025}
}
R2 v1 2026-07-01T02:32:52.645Z