all AI news
Dynamic Memory Compression: Retrofitting LLMs for Accelerated Inference
March 15, 2024, 4:48 a.m. | Piotr Nawrot, Adrian {\L}a\'ncucki, Marcin Chochowski, David Tarjan, Edoardo M. Ponti
cs.CL updates on arXiv.org arxiv.org
Abstract: Transformers have emerged as the backbone of large language models (LLMs). However, generation remains inefficient due to the need to store in memory a cache of key-value representations for past tokens, whose size scales linearly with the input sequence length and batch size. As a solution, we propose Dynamic Memory Compression (DMC), a method for on-line key-value cache compression at inference time. Most importantly, the model learns to apply different compression rates in different heads …
abstract arxiv cache compression cs.cl dynamic however inference key language language models large language large language models llms memory solution store tokens transformers type value
More from arxiv.org / cs.CL updates on arXiv.org
Jobs in AI, ML, Big Data
Artificial Intelligence – Bioinformatic Expert
@ University of Texas Medical Branch | Galveston, TX
Lead Developer (AI)
@ Cere Network | San Francisco, US
Research Engineer
@ Allora Labs | Remote
Ecosystem Manager
@ Allora Labs | Remote
Founding AI Engineer, Agents
@ Occam AI | New York
AI Engineer Intern, Agents
@ Occam AI | US