mirror of
https://github.com/dogkeeper886/ollama37.git
synced 2025-12-14 17:57:06 +00:00
warning F16 uses significantly more memory than quantized model so the standard requires don't apply.
2.3 KiB
2.3 KiB