I Pruned Layers from a 30B Code Model and It Got 123% Better at Benchmarks (gabeortiz.net)

3 points by egonschiele 3mo ago ↗ HN

0 comments

[ 1.7 ms ] story [ 9.3 ms ] thread

No comments yet.

A showcase of shard-db. Data from the Hacker News API. Refreshed every 15 minutes.

Source · Live DB stats