KServe deploy

module-5/.gitignore

@@ -1 +1,2 @@
 .lock-file
+seldon-core

module-5/README.md

@@ -102,12 +102,38 @@ make run_pytriton
 
 # KServe 
 
-Install 
+Install KServe
 
 ```
 curl -s "https://raw.githubusercontent.com/kserve/kserve/release-0.13/hack/quick_install.sh" | bash
 ```
 
+Deploy custom model
+
+```
+kubectl create -f ./k8s/kserve-inferenceserver.yaml
+```
+
+Call API 
+
+```
+kubectl get inferenceservices custom-model
+kubectl get svc --namespace istio-system
+kubectl port-forward --namespace istio-system svc/istio-ingressgateway 8080:80
+
+curl -v -H "Host: custom-model.default.example.com" -H "Content-Type: application/json" "http://localhost:8080/v1/models/custom-model:predict" -d @data-samples/iris-input.json
+
+```
+
+
+# Seldon V2 
+
+```
+git clone https://github.com/SeldonIO/seldon-core --branch=v2
+```
+
+
+
 ## IRIS
 
 ```

module-5/k8s/app-triton.yaml

@@ -0,0 +1,37 @@
+---
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: app-triton
+spec:
+  replicas: 2
+  selector:
+    matchLabels:
+      app: app-triton
+  template:
+    metadata:
+      labels:
+        app: app-triton
+    spec:
+      containers:
+        - name: app-triton
+          image: ghcr.io/kyryl-opens-ml/app-triton:latest
+          env:
+          - name: WANDB_API_KEY
+            valueFrom:
+              secretKeyRef:
+                name: wandb
+                key: WANDB_API_KEY
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: app-triton
+  labels:
+    app: app-triton
+spec:
+  ports:
+  - port: 8080
+    protocol: TCP
+  selector:
+    app: app-triton

module-5/k8s/kserve-inferenceserver.yaml

@@ -0,0 +1,15 @@
+apiVersion: serving.kserve.io/v1beta1
+kind: InferenceService
+metadata:
+  name: custom-model
+spec:
+  predictor:
+    containers:
+      - name: kserve-container
+        image: ghcr.io/kyryl-opens-ml/app-kserve:latest
+        env:
+        - name: WANDB_API_KEY
+          valueFrom:
+            secretKeyRef:
+              name: wandb
+              key: WANDB_API_KEY

module-5/serving/kserve_api.py

@@ -15,8 +15,11 @@ def load(self):
         self.ready = True
 
     def predict(self, payload: Dict, headers: Dict[str, str] = None) -> Dict:
-        json_payload = json.loads(payload.decode("utf-8"))
-        instances = json_payload["instances"]
+        print(payload)
+        print(type(payload))
+        # json_payload = json.loads(payload.decode("utf-8"))
+        # instances = json_payload["instances"]
+        instances = payload["instances"]
         predictions = self.predictor.predict(instances)
         return {"predictions": predictions.tolist()}