Is there any way to use CLI deploy model to kubernetes

[crane136]

I want to use bentoml apply command to deplyment model to kubernetes. Also the deployments can be managed by BENTOML UI Tools

I got some problems When I use the CLI: ( Is there any way to use CLI deploy model to kubernetes?)

iris-classifier

apiVersion: v1
kind: Service
metadata:
labels:
app: iris-classifier
name: iris-classifier
spec:
ports:
- name: predict
port: 5000
targetPort: 5000
selector:
app: iris-classifier
type: LoadBalancer

---

apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: iris-classifier
name: iris-classifier
spec:
selector:
matchLabels:
app: iris-classifier
template:
metadata:
labels:
app: iris-classifier
spec:
containers:
- image: iris-classifier:20210226105119_51D3D8
imagePullPolicy: IfNotPresent
name: iris-classifier
ports:
- containerPort: 5000

pool-100-8-12-202:models crane$ bentoml deployment apply -f iris-classifier.yaml
Usage: bentoml deployment apply [OPTIONS]

Error: Invalid value for '-f' / '--file': Input value is not recognizable yaml file or yaml content

Try another:

apiVersion: serving.kubeflow.org/v1alpha2
kind: InferenceService
metadata:
labels:
controller-tools.k8s.io: "1.0"
name: iris-classifier
spec:
default:
predictor:
custom:
container:
image: iris-classifier:20210226105119_51D3D8
ports:
- containerPort: 5000

pool-100-8-12-202:models crane$ bentoml deployment apply -f bentoml.yaml
Error: deployment apply failed: Platform "None" is not supported in the current version of BentoML

[yubozhao]

@crane136 BentoML doesn't provide Kubernetes deployment at the moment. This feature is on our road map and we are actively designing and working on it.

I would love to get your input on what you are looking for. Can you tell me more about your use-cases and current setup?

I am also curious about the Kubeflow inference service you had in your "Try another" section. Do you use Kubeflow in your workflow? What part of Kubeflow system you guys are using?