test qwen

clusters/k3s-dgx/apps/huihui-granite-inference.yaml

@@ -1,7 +1,7 @@
-apiVersion: serving.kserve.io/v1beta1
-kind: InferenceService
+apiVersion: "serving.kserve.io/v1beta1"
+kind: "InferenceService"
 metadata:
-  name: huihui-granite
+  name: "qwen-llm"
   namespace: kserve
 spec:
   predictor:
@@ -9,14 +9,14 @@ spec:
       modelFormat:
         name: huggingface
       args:
-        - --model_name=huihui-granite
-      storageUri: "hf://huihui-ai/Huihui-granite-4.1-30b-abliterated"
+        - --model_name=qwen
+      storageUri: "hf://Qwen/Qwen2.5-0.5B-Instruct"
       resources:
         limits:
-          cpu: "4"
-          memory: 16Gi
+          cpu: "2"
+          memory: 6Gi
           nvidia.com/gpu: "1"
         requests:
-          cpu: "2"
-          memory: 8Gi
-          nvidia.com/gpu: "1"
+          cpu: "1"
+          memory: 4Gi
+          nvidia.com/gpu: "1"