diff --git a/start_all.sh b/start_all.sh index 730e598c..cf79f1cf 100755 --- a/start_all.sh +++ b/start_all.sh @@ -234,7 +234,7 @@ start_services() { --port ${INFERENCE_PORT} \ --gpu-layers 36 \ --host 0.0.0.0 \ - -c 8192 \ + -c 16384 \ --jinja \ --temp 0 \ --top-k 1 \