ollama: elimina GPU, fija imagen 0.20.7, reduce a qwen2.5:3b

- Imagen: ollama/ollama:latest → ollama/ollama:0.20.7
- Elimina NVIDIA_VISIBLE_DEVICES, NVIDIA_DRIVER_CAPABILITIES
- Elimina nvidia.com/gpu: "1" de resources limits
- Reduce memory: 8/20Gi → 4/8Gi (solo CPU, modelo 3b)
- Startup: auto-pull cambiado a qwen2.5:3b

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

ollama/ollama.yaml

@@ -42,7 +42,7 @@ spec:
         kubernetes.io/hostname: chemavx-k8
       containers:
         - name: ollama
-          image: ollama/ollama:latest
+          image: ollama/ollama:0.20.7
           command: ["/usr/bin/bash", "-c"]
           args:
             - |
@@ -53,8 +53,7 @@ spec:
                 sleep 2
               done
               echo "Pulling llama3.1:8b..."
-              ollama pull llama3.1:8b
+              ollama pull qwen2.5:3b
-              ollama rm llama3.2:3b 2>/dev/null || true
               echo "Model ready."
               wait $SERVE_PID
           ports:
@@ -68,17 +67,12 @@ spec:
               value: "8192"
             - name: OLLAMA_KEEP_ALIVE
               value: "-1"
-            - name: NVIDIA_VISIBLE_DEVICES
-              value: all
-            - name: NVIDIA_DRIVER_CAPABILITIES
-              value: compute,utility
           resources:
             requests:
-              memory: 8Gi
+              memory: 4Gi
               cpu: "500m"
             limits:
-              memory: 20Gi
+              memory: 8Gi
-              nvidia.com/gpu: "1"
           volumeMounts:
             - name: ollama-data
               mountPath: /root/.ollama