from Hacker News

Benchmarking LMs for Dedupe

by thatjoeoverthr on 3/18/25, 10:05 PM with 1 comments

  • by thatjoeoverthr on 3/18/25, 10:05 PM

    I needed to understand my options for semantic deduplication, so I went through a process of benchmarking several LMs and wrote it all up here. I hope it's useful to someone!