add ollama back to artemision
All checks were successful
Check flake.lock / Check health of `flake.lock` (pull_request) Successful in 6s
Check Nix flake / Perform Nix flake checks (pull_request) Successful in 15m11s

This commit is contained in:
2026-03-26 21:40:12 -04:00
parent b5e45398d8
commit 802bf1ca4c
4 changed files with 45 additions and 27 deletions

View File

@@ -25,11 +25,12 @@ in
"nemotron-cascade-2:30b"
"magistral"
"devstral-small-2"
"starcoder2:15b"
];
models = vars.primary_ollama;
environmentVariables = {
FLASH_ATTENTION = "1";
OLLAMA_KV_CACHE_TYPE = "q8_0";
OLLAMA_KV_CACHE_TYPE = "q4_0";
# Ollama memory configuration
OLLAMA_MAX_LOADED_MODELS = "3";
OLLAMA_MAX_QUEUE = "512";