KServe

https://kserve.github.io/website/0.10/modelserving/control_plane/

https://kserve.github.io/website/0.10/modelserving/data_plane/data_plane/

설치

• https://github.com/kserve/kserve/releases

helm upgrade kserve-crd kserve-crd-v0.11.2.tgz \
  --install \
  --history-max 5

helm show values kserve-v0.11.2.tgz \
  > kserve-values.yaml

kserve-values.yaml

kserve:
  controller:
    gateway:
      domain: <domain>
      localGateway:
        # InferenceService를 생성하면
        # gateway로 들어오는 요청을 gatewayService:80으로 전달하는
        # VirtualService를 생성합니다
        gateway: <gatewayNamespace>/<gatewayName>
        gatewayService: <ingressGatewayServiceHost>
      ingressGateway:
        gateway: <gatewayNamespace>/<gatewayName>
        gatewayService: <ingressGatewayServiceHost>
  modelmesh:
    enabled: false

helm template kserve kserve-v0.11.2.tgz \
  -n kserve \
  -f kserve-values.yaml \
  > kserve.yaml

helm upgrade kserve kserve-v0.11.2.tgz \
  --install \
  --history-max 5 \
  -n kserve \
  -f kserve-values.yaml

Test

kubectl create namespace kserve-test

apiVersion: serving.kserve.io/v1beta1
kind: InferenceService
metadata:
  name: sklearn-iris
  namespace: kserve-test
spec:
  predictor:
    model:
      modelFormat:
        name: sklearn
      storageUri: gs://kfserving-examples/models/sklearn/1.0/model
      args: ["--enable_docs_url=True"]

curl -v \
  -H "Host: sklearn-iris.kserve-test.svc.cluster.local" \
  http://localhost:8081/v1/models/sklearn-iris:predict \
  -d '{"instances": [[6.8,  2.8,  4.8,  1.4],[6.0,  3.4,  4.5,  1.6]]}'