diff --git a/clusters/k3s-dgx/nim-service/qwen.yaml b/clusters/k3s-dgx/nim-service/qwen.yaml index e09e982..702d447 100644 --- a/clusters/k3s-dgx/nim-service/qwen.yaml +++ b/clusters/k3s-dgx/nim-service/qwen.yaml @@ -87,6 +87,8 @@ spec: authSecret: hf-api-secret storage: + sharedMemorySizeLimit: + sizeLimit: 64Gi nimCache: name: qwen36-27b-fp8 replicas: 1