Show HN: I finetuned 1.5B Qwen to near GPT-4o level bash generation perf
By GodelNumbering · 2026-10-05 · 4 points · 0 comments
https://dirac.run/posts/easycommand
Purely a hobby side project to see how far I can push a really small model, using (mostly) automated training pipelines Original mention: https://news.ycombinator.com/item?id=49869735 There were a bunch of requests to release it. Full synthetic data: https:/&…
Open the full discussion on BetterNews