The weights fit. The inference workload didn't. 0 ▲ Siddhant Khare 1 day ago · Tech · hide · 0 comments A 16,081,490,933-byte model hit a clean-boot Metal OOM at 256 prompt tokens on a 24 GiB M5 Pro. Stage checkpoints show why file size was never a fit proof. No comments yet. Log in to reply on the Fediverse. Comments will appear here.
