mirror of
https://github.com/likelovewant/ollama-for-amd.git
synced 2025-12-23 15:08:27 +00:00
In order to iteratively find the best memory allocation, we need to be able to free backend memory so we can try again.
36 KiB
36 KiB