all AI news
Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve
March 5, 2024, 2:42 p.m. | Amey Agrawal, Nitin Kedia, Ashish Panwar, Jayashree Mohan, Nipun Kwatra, Bhargav S. Gulavani, Alexey Tumanov, Ramachandran Ramjee
cs.LG updates on arXiv.org arxiv.org
Abstract: Each LLM serving request goes through two phases. The first is prefill which processes the entire input prompt to produce one output token and the second is decode which generates the rest of output tokens, one-at-a-time. Prefill iterations have high latency but saturate GPU compute due to parallel processing of the input prompt. In contrast, decode iterations have low latency but also low compute utilization because a decode iteration processes only a single token per …
abstract arxiv compute cs.dc cs.lg decode gpu inference latency llm processes prompt request rest serve through token tokens type
More from arxiv.org / cs.LG updates on arXiv.org
Jobs in AI, ML, Big Data
Data Architect
@ University of Texas at Austin | Austin, TX
Data ETL Engineer
@ University of Texas at Austin | Austin, TX
Lead GNSS Data Scientist
@ Lurra Systems | Melbourne
Senior Machine Learning Engineer (MLOps)
@ Promaton | Remote, Europe
Data Engineer - AWS
@ 3Pillar Global | Costa Rica
Cost Controller/ Data Analyst - India
@ John Cockerill | Mumbai, India, India, India