1 comment

[ 3.7 ms ] story [ 15.0 ms ] thread
We study how to approximate the famous shortest-job-first scheduling in LLM inference!