enable gpu acceleration for openedai-speech

kustomize/bases/ollama/ollama-deployment.yaml

@@ -43,7 +43,7 @@ spec:
             limits:
               memory: 4Gi
               cpu: 3000m
-              nvidia.com/gpu: "2"
+              nvidia.com/gpu: "1"
               # amd.com/gpu: "1"
           ports:
             - name: http

kustomize/bases/openedai-speech/openedai-speech-deployment.yaml

@@ -26,9 +26,13 @@ spec:
                     operator: In
                     values:
                       - amd64
+      runtimeClassName: nvidia
       containers:
         - name: openedai-speech
           image: ghcr.io/matatonic/openedai-speech
+          env:
+            - name: EXTRA_ARGS
+              value: --preload xtts_v2.0.2
           ports:
             - name: http
               containerPort: 8000
@@ -38,7 +42,7 @@ spec:
               memory: 1Gi
             limits:
               cpu: 1000m
-              memory: 1Gi
+              nvidia.com/gpu: "1"
           volumeMounts:
             - name: server-voices-pv
               mountPath: /app/voices