gpu offload in llama.cpp

2025-07-23 19:51:26 +02:00
parent 5fb2bcfc7e
commit d53db88fd2
2 changed files with 3 additions and 0 deletions
--- a/apps/llama/deployment.yaml
+++ b/apps/llama/deployment.yaml
@@ -18,6 +18,8 @@ spec:
        - name: llama-swap
          image: ghcr.io/mostlygeek/llama-swap:v139-vulkan-b5957
          imagePullPolicy: IfNotPresent
+          command:
+            - /app/llama-swap
          args:
            - --config=/config/config.yaml
            - --watch-config