straylight: raise OpenCode and llama-server timeouts to 2h
Local prefill and reasoning/coding turns often sit idle past OpenCode's 5–10 minute defaults. Align llama-server --timeout with that window.
This commit is contained in:
@@ -26,6 +26,12 @@ let
|
||||
options = {
|
||||
baseURL = llamaBase;
|
||||
apiKey = "none";
|
||||
# Local prefill + Ornith <think> / Laguna coding turns routinely sit
|
||||
# quiet for >10 minutes. OpenCode defaults are 300000ms (request,
|
||||
# headers, and inter-chunk). Milliseconds; false disables.
|
||||
timeout = 7200000;
|
||||
headerTimeout = 7200000;
|
||||
chunkTimeout = 7200000;
|
||||
};
|
||||
models = {
|
||||
"laguna-s-2.1" = text "Laguna S 2.1";
|
||||
|
||||
@@ -187,7 +187,8 @@ in
|
||||
--host 0.0.0.0 \
|
||||
--port 11434 \
|
||||
--models-preset ${llamaModelsIni} \
|
||||
--models-max 2
|
||||
--models-max 2 \
|
||||
--timeout 7200
|
||||
'';
|
||||
User = "llama-server";
|
||||
Group = "llama-server";
|
||||
|
||||
Reference in New Issue
Block a user