mirror of
https://github.com/cloudstack-llc/ollama.git
synced 2026-07-19 22:53:30 -04:00
c3d321d405
This workaround logic in llama.cpp is causing crashes for users with less system memory than VRAM.