mirror of
https://github.com/likelovewant/ollama-for-amd.git
synced 2025-12-22 06:43:57 +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.
12 KiB
12 KiB