update kubeflow dip-catalog
This commit is contained in:
@@ -0,0 +1,56 @@
|
||||
KSERVE_VERSION ?= 0.12.1
|
||||
MODELS_WEBAPP_VERSION ?= 0.8.1
|
||||
MODELS_WEBAPP_RELEASE_VERSION := $(shell echo ${MODELS_WEBAPP_VERSION} | cut -d "." -f1-2)
|
||||
|
||||
.PHONY: upgrade-kserve-manifests
|
||||
upgrade-kserve-manifests: clean-kserve-manifests
|
||||
curl -sSL 'https://github.com/kserve/kserve/releases/download/v$(KSERVE_VERSION)/kserve_kubeflow.yaml' -o 'kserve/kserve_kubeflow.yaml'
|
||||
curl -sSL 'https://github.com/kserve/kserve/releases/download/v$(KSERVE_VERSION)/kserve-cluster-resources.yaml' -o 'kserve/kserve-cluster-resources.yaml'
|
||||
curl -sSL 'https://github.com/kserve/kserve/releases/download/v$(KSERVE_VERSION)/kserve.yaml' -o 'kserve/kserve.yaml'
|
||||
|
||||
.PHONY: clean-kserve-manifests
|
||||
clean-kserve-manifests:
|
||||
cd kserve && rm -f kserve.yaml kserve-cluster-resources.yaml kserve_kubeflow.yaml
|
||||
|
||||
.PHONY: install-kserve
|
||||
install-kserve:
|
||||
# Create kubeflow namespace if not exists
|
||||
kubectl create namespace kubeflow || echo "kubeflow namespace exists"
|
||||
kubectl apply -k kserve
|
||||
|
||||
.PHONY: uninstall-kserve
|
||||
uninstall-kserve:
|
||||
kubectl delete -k kserve
|
||||
|
||||
.PHONY: test-kserve
|
||||
test-kserve:
|
||||
cd tests && pytest .
|
||||
|
||||
.PHONY: clean-models-webapp-manifests
|
||||
clean-models-webapp-manifests:
|
||||
rm -rf models-web-app
|
||||
|
||||
.PHONY: upgrade-models-webapp-manifests
|
||||
upgrade-models-webapp-manifests: clean-models-webapp-manifests
|
||||
git clone https://github.com/kserve/models-web-app.git kserve-models-web-app-src
|
||||
cd kserve-models-web-app-src && git checkout release-${MODELS_WEBAPP_RELEASE_VERSION}
|
||||
cp -r kserve-models-web-app-src/config ./models-web-app
|
||||
# cleanup
|
||||
rm -rf kserve-models-web-app-src
|
||||
|
||||
.PHONY: install-models-web-app
|
||||
install-models-webapp:
|
||||
# Create kubeflow namespace if not exists
|
||||
kubectl create namespace kubeflow || echo "kubeflow namespace exists"
|
||||
kubectl apply -k models-web-app/overlays/kubeflow
|
||||
|
||||
.PHONY: uninstall-models-web-app
|
||||
uninstall-models-webapp:
|
||||
kubectl delete -k models-web-app/overlays/kubeflow
|
||||
|
||||
.PHONY: test-models-webapp
|
||||
test-models-webapp: install-models-webapp
|
||||
kubectl wait --for=condition=Available --timeout=300s -n kubeflow deployment/kserve-models-web-app
|
||||
@echo "Test Passed"
|
||||
@echo "cleaning up ..."
|
||||
kubectl delete -k models-web-app/overlays/kubeflow
|
||||
@@ -0,0 +1,2 @@
|
||||
approvers:
|
||||
- yuzisun
|
||||
@@ -0,0 +1,120 @@
|
||||
# KServe
|
||||
[](https://goreportcard.com/report/github.com/kserve/kserve)
|
||||
[](https://bestpractices.coreinfrastructure.org/projects/6643)
|
||||
[](https://github.com/kserve/kserve/releases)
|
||||
[](https://github.com/kserve/kserve/blob/master/LICENSE)
|
||||
|
||||
KServe provides a Kubernetes [Custom Resource Definition](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) for serving predictive and generative machine learning (ML) models. It aims to solve production model serving use cases by providing high abstraction interfaces for Tensorflow, XGBoost, ScikitLearn, PyTorch, Huggingface Transformer/LLM models using standardized data plane protocols.
|
||||
|
||||
It encapsulates the complexity of autoscaling, networking, health checking, and server configuration to bring cutting edge serving features like GPU Autoscaling, Scale to Zero, and Canary Rollouts to your ML deployments. It enables a simple, pluggable, and complete story for Production ML Serving including prediction, pre-processing, post-processing and explainability. KServe is being [used across various organizations.](https://kserve.github.io/website/master/community/adopters/)
|
||||
|
||||
For more details, visit the [KServe website](https://kserve.github.io/website/).
|
||||
|
||||

|
||||
|
||||
*[KFServing has been rebranded to KServe since v0.7](https://blog.kubeflow.org/release/official/2021/09/27/kfserving-transition.html).*
|
||||
|
||||
## Why KServe?
|
||||
- KServe is a standard, cloud agnostic **Model Inference Platform** for serving predictive and generative AI models on Kubernetes, built for highly scalable use cases.
|
||||
- Provides performant, **standardized inference protocol** across ML frameworks including OpenAI specification for generative models.
|
||||
- Support modern **serverless inference workload** with **request based autoscaling including scale-to-zero** on **CPU and GPU**.
|
||||
- Provides **high scalability, density packing and intelligent routing** using **ModelMesh**.
|
||||
- **Simple and pluggable production serving** for **inference**, **pre/post processing**, **monitoring** and **explainability**.
|
||||
- Advanced deployments for **canary rollout**, **pipeline**, **ensembles** with **InferenceGraph**.
|
||||
|
||||
|
||||
## :hammer_and_wrench: Installation
|
||||
|
||||
### Standalone Installation
|
||||
- **[Serverless Installation](https://kserve.github.io/website/master/admin/serverless/serverless/)**: KServe by default installs Knative for **serverless deployment** for InferenceService.
|
||||
- **[Raw Deployment Installation](https://kserve.github.io/website/master/admin/kubernetes_deployment)**: Compared to Serverless Installation, this is a more **lightweight** installation. However, this option does not support canary deployment and request based autoscaling with scale-to-zero.
|
||||
- **[ModelMesh Installation](https://kserve.github.io/website/master/admin/modelmesh/)**: You can optionally install ModelMesh to enable **high-scale**, **high-density** and **frequently-changing model serving** use cases.
|
||||
- **[Quick Installation](https://kserve.github.io/website/master/get_started/)**: Install KServe on your local machine.
|
||||
|
||||
### Kubeflow Installation
|
||||
KServe is an important addon component of Kubeflow, please learn more from the [Kubeflow KServe documentation](https://www.kubeflow.org/docs/external-add-ons/kserve/kserve) and follow [KServe with Kubeflow on AWS](https://awslabs.github.io/kubeflow-manifests/main/docs/component-guides/kserve) to learn how to use KServe on AWS.
|
||||
|
||||
## :hammer_and_pick: Models Web App
|
||||
The Models web app is responsible for allowing the user to manipulate the Model Servers in their Kubeflow cluster. To achieve this it provides a user friendly way to handle the lifecycle of InferenceService CRs.
|
||||
Please follow the [Kserve Models UI documentation](https://www.kubeflow.org/docs/external-add-ons/kserve/webapp/) for more information.
|
||||
|
||||
## :rocket: Upgrading
|
||||
For upgrading see [UPGRADE.md](UPGRADE.md)
|
||||
|
||||
## :microscope: Testing
|
||||
### Testing Kserve
|
||||
#### Prerequisite
|
||||
|
||||
1. Install Python >= 3.8
|
||||
2. Install requirements
|
||||
```sh
|
||||
pip install -r tests/requirements.txt
|
||||
```
|
||||
3. Create kubeflow namespace
|
||||
```sh
|
||||
kubectl apply -k ../../common/kubeflow-namespace/base
|
||||
```
|
||||
4. Install cert manager
|
||||
```sh
|
||||
kubectl apply -k ../../common/cert-manager/cert-manager/base
|
||||
kubectl apply -k ../../common/cert-manager/kubeflow-issuer/base
|
||||
```
|
||||
5. Install Istio
|
||||
```sh
|
||||
kubectl apply -k ../../common/istio-1-22/istio-crds/base
|
||||
kubectl apply -k ../../common/istio-1-22/istio-namespace/base
|
||||
kubectl apply -k ../../common/istio-1-22/istio-install/base
|
||||
```
|
||||
6. Install knative
|
||||
```sh
|
||||
kubectl apply -k ../../common/knative/knative-serving/overlays/gateways
|
||||
kubectl apply -k ../../common/istio-1-22/cluster-local-gateway/base
|
||||
kubectl apply -k ../../common/istio-1-22/kubeflow-istio-resources/base
|
||||
```
|
||||
7. Install kserve
|
||||
```sh
|
||||
make install-kserve
|
||||
```
|
||||
> **_NOTE:_** If resource/crd installation fails please re-run the commands.
|
||||
|
||||
#### Steps
|
||||
1. Create test namespace
|
||||
```sh
|
||||
kubectl create ns kserve-test
|
||||
```
|
||||
2. Configure domain name
|
||||
```sh
|
||||
kubectl patch cm config-domain --patch '{"data":{"example.com":""}}' -n knative-serving
|
||||
```
|
||||
|
||||
3. Port forward
|
||||
```sh
|
||||
# start a new terminal and run
|
||||
INGRESS_GATEWAY_SERVICE=$(kubectl get svc --namespace istio-system --selector="app=istio-ingressgateway" --output jsonpath='{.items[0].metadata.name}')
|
||||
kubectl port-forward --namespace istio-system svc/${INGRESS_GATEWAY_SERVICE} 8080:80
|
||||
```
|
||||
4. Run test
|
||||
```sh
|
||||
export KSERVE_INGRESS_HOST_PORT='localhost:8080'
|
||||
make test-kserve
|
||||
|
||||
### Testing Models WebApp
|
||||
#### Prerequisite
|
||||
1. Running kubernetes cluster
|
||||
2. `kubectl` configured to talk to the desired cluster.
|
||||
|
||||
#### Steps
|
||||
1. Run the test
|
||||
```sh
|
||||
make test-models-webapp
|
||||
```
|
||||
## :flight_departure: [Create your first InferenceService](https://kserve.github.io/website/master/get_started/first_isvc)
|
||||
|
||||
## :blue_book: [InferenceService API Reference](https://kserve.github.io/website/master/reference/api/)
|
||||
|
||||
## :handshake: [Adopters](https://kserve.github.io/website/master/community/adopters/)
|
||||
|
||||
## :books: Learn More
|
||||
To learn more about KServe, how to use various supported features, and how to participate in the KServe community,
|
||||
please follow the [KServe website documentation](https://kserve.github.io/website).
|
||||
Additionally, we have compiled a list of [presentations and demos](https://kserve.github.io/website/master/community/presentations/) to dive through various details.
|
||||
@@ -0,0 +1,68 @@
|
||||
# Upgrade Documentation
|
||||
|
||||
## Upgrade Kserve
|
||||
|
||||
### Prerequisites
|
||||
|
||||
- Running Kubernetes cluster with kubeflow installed.
|
||||
- `kubectl` configured to talk to the desired cluster.
|
||||
- `curl`
|
||||
|
||||
> **_NOTE:_** This documentation assumes that you are running the commands in linux.
|
||||
If you are using another OS, please make sure to update the Makefile commands.
|
||||
|
||||
### To update the kserve manifests to specific version follow the below instructions.
|
||||
|
||||
1. Set the desired version to upgrade.
|
||||
|
||||
```sh
|
||||
export KSERVE_VERSION=0.12.1
|
||||
```
|
||||
|
||||
2. Rebuild the manifests.
|
||||
|
||||
```sh
|
||||
make upgrade-kserve-manifests
|
||||
```
|
||||
|
||||
3. Install the updated manifests.
|
||||
```sh
|
||||
make install-kserve
|
||||
```
|
||||
> **_NOTE:_** If resource/crd installation fails please re-run the commands.
|
||||
|
||||
### Testing
|
||||
For testing refer [kserve readme](README.md#testing-kserve).
|
||||
|
||||
## Upgrade Models Webapp
|
||||
### Prerequisites
|
||||
|
||||
- Running Kubernetes cluster with kubeflow installed.
|
||||
- `kubectl` configured to talk to the desired cluster.
|
||||
- `git`
|
||||
|
||||
> **_NOTE:_** This documentation assumes that you are running the commands in linux.
|
||||
If you are using another OS, please make sure to update the Makefile commands.
|
||||
|
||||
### To update the kserve manifests to specific version follow the below instructions.
|
||||
|
||||
1. Set the desired version to upgrade.
|
||||
|
||||
```sh
|
||||
export MODELS_WEBAPP_VERSION=0.8.1
|
||||
```
|
||||
|
||||
2. Rebuild the manifests.
|
||||
|
||||
```sh
|
||||
make upgrade-models-webapp-manifests
|
||||
```
|
||||
|
||||
3. Install the updated manifests.
|
||||
```sh
|
||||
make install-models-webapp
|
||||
```
|
||||
> **_NOTE:_** If resource/crd installation fails please re-run the commands.
|
||||
|
||||
### Testing
|
||||
For testing refer [kserve readme](README.md#testing-models-webapp).
|
||||
Binary file not shown.
|
After Width: | Height: | Size: 462 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 99 KiB |
@@ -0,0 +1,89 @@
|
||||
aggregationRule:
|
||||
clusterRoleSelectors:
|
||||
- matchLabels:
|
||||
rbac.authorization.kubeflow.org/aggregate-to-kubeflow-kserve-admin: "true"
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRole
|
||||
metadata:
|
||||
labels:
|
||||
app: kserve
|
||||
app.kubernetes.io/name: kserve
|
||||
rbac.authorization.kubeflow.org/aggregate-to-kubeflow-admin: "true"
|
||||
name: kubeflow-kserve-admin
|
||||
rules: []
|
||||
---
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRole
|
||||
metadata:
|
||||
labels:
|
||||
app: kserve
|
||||
app.kubernetes.io/name: kserve
|
||||
rbac.authorization.kubeflow.org/aggregate-to-kubeflow-edit: "true"
|
||||
rbac.authorization.kubeflow.org/aggregate-to-kubeflow-kserve-admin: "true"
|
||||
name: kubeflow-kserve-edit
|
||||
rules:
|
||||
- apiGroups:
|
||||
- serving.kserve.io
|
||||
resources:
|
||||
- inferenceservices
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
- watch
|
||||
- create
|
||||
- delete
|
||||
- deletecollection
|
||||
- patch
|
||||
- update
|
||||
- apiGroups:
|
||||
- serving.knative.dev
|
||||
resources:
|
||||
- services
|
||||
- services/status
|
||||
- routes
|
||||
- routes/status
|
||||
- configurations
|
||||
- configurations/status
|
||||
- revisions
|
||||
- revisions/status
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
- watch
|
||||
- create
|
||||
- delete
|
||||
- deletecollection
|
||||
- patch
|
||||
- update
|
||||
---
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRole
|
||||
metadata:
|
||||
labels:
|
||||
app: kserve
|
||||
app.kubernetes.io/name: kserve
|
||||
rbac.authorization.kubeflow.org/aggregate-to-kubeflow-view: "true"
|
||||
name: kubeflow-kserve-view
|
||||
rules:
|
||||
- apiGroups:
|
||||
- serving.kserve.io
|
||||
resources:
|
||||
- inferenceservices
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
- watch
|
||||
- apiGroups:
|
||||
- serving.knative.dev
|
||||
resources:
|
||||
- services
|
||||
- services/status
|
||||
- routes
|
||||
- routes/status
|
||||
- configurations
|
||||
- configurations/status
|
||||
- revisions
|
||||
- revisions/status
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
@@ -0,0 +1,391 @@
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-huggingfaceserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
image: kserve/huggingfaceserver:v0.13.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v2
|
||||
- v1
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: huggingface
|
||||
priority: 1
|
||||
version: "1"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-lgbserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
- --model_dir=/mnt/models
|
||||
- --http_port=8080
|
||||
- --nthread=1
|
||||
image: kserve/lgbserver:v0.13.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: lightgbm
|
||||
priority: 1
|
||||
version: "3"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-mlserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- env:
|
||||
- name: MLSERVER_MODEL_IMPLEMENTATION
|
||||
value: '{{.Labels.modelClass}}'
|
||||
- name: MLSERVER_HTTP_PORT
|
||||
value: "8080"
|
||||
- name: MLSERVER_GRPC_PORT
|
||||
value: "9000"
|
||||
- name: MODELS_DIR
|
||||
value: /mnt/models
|
||||
image: docker.io/seldonio/mlserver:1.3.2
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: sklearn
|
||||
priority: 2
|
||||
version: "0"
|
||||
- autoSelect: true
|
||||
name: sklearn
|
||||
priority: 2
|
||||
version: "1"
|
||||
- autoSelect: true
|
||||
name: xgboost
|
||||
priority: 2
|
||||
version: "1"
|
||||
- autoSelect: true
|
||||
name: lightgbm
|
||||
priority: 2
|
||||
version: "3"
|
||||
- autoSelect: true
|
||||
name: mlflow
|
||||
priority: 1
|
||||
version: "1"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-paddleserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
- --model_dir=/mnt/models
|
||||
- --http_port=8080
|
||||
image: kserve/paddleserver:v0.13.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: paddle
|
||||
priority: 1
|
||||
version: "2"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-pmmlserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
- --model_dir=/mnt/models
|
||||
- --http_port=8080
|
||||
image: kserve/pmmlserver:v0.13.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: pmml
|
||||
priority: 1
|
||||
version: "3"
|
||||
- autoSelect: true
|
||||
name: pmml
|
||||
priority: 1
|
||||
version: "4"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-sklearnserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
- --model_dir=/mnt/models
|
||||
- --http_port=8080
|
||||
image: kserve/sklearnserver:v0.13.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: sklearn
|
||||
priority: 1
|
||||
version: "1"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-tensorflow-serving
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
- --port=9000
|
||||
- --rest_api_port=8080
|
||||
- --model_base_path=/mnt/models
|
||||
- --rest_api_timeout_in_ms=60000
|
||||
command:
|
||||
- /usr/bin/tensorflow_model_server
|
||||
image: tensorflow/serving:2.6.2
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- grpc-v1
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: tensorflow
|
||||
priority: 2
|
||||
version: "1"
|
||||
- autoSelect: true
|
||||
name: tensorflow
|
||||
priority: 2
|
||||
version: "2"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-torchserve
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8082"
|
||||
containers:
|
||||
- args:
|
||||
- torchserve
|
||||
- --start
|
||||
- --model-store=/mnt/models/model-store
|
||||
- --ts-config=/mnt/models/config/config.properties
|
||||
env:
|
||||
- name: TS_SERVICE_ENVELOPE
|
||||
value: '{{.Labels.serviceEnvelope}}'
|
||||
image: pytorch/torchserve-kfs:0.9.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- v2
|
||||
- grpc-v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: pytorch
|
||||
priority: 2
|
||||
version: "1"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-tritonserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8002"
|
||||
containers:
|
||||
- args:
|
||||
- tritonserver
|
||||
- --model-store=/mnt/models
|
||||
- --grpc-port=9000
|
||||
- --http-port=8080
|
||||
- --allow-grpc=true
|
||||
- --allow-http=true
|
||||
image: nvcr.io/nvidia/tritonserver:23.05-py3
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v2
|
||||
- grpc-v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: tensorrt
|
||||
priority: 1
|
||||
version: "8"
|
||||
- autoSelect: true
|
||||
name: tensorflow
|
||||
priority: 1
|
||||
version: "1"
|
||||
- autoSelect: true
|
||||
name: tensorflow
|
||||
priority: 1
|
||||
version: "2"
|
||||
- autoSelect: true
|
||||
name: onnx
|
||||
priority: 1
|
||||
version: "1"
|
||||
- name: pytorch
|
||||
version: "1"
|
||||
- autoSelect: true
|
||||
name: triton
|
||||
priority: 1
|
||||
version: "2"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterServingRuntime
|
||||
metadata:
|
||||
name: kserve-xgbserver
|
||||
spec:
|
||||
annotations:
|
||||
prometheus.kserve.io/path: /metrics
|
||||
prometheus.kserve.io/port: "8080"
|
||||
containers:
|
||||
- args:
|
||||
- --model_name={{.Name}}
|
||||
- --model_dir=/mnt/models
|
||||
- --http_port=8080
|
||||
- --nthread=1
|
||||
image: kserve/xgbserver:v0.13.0
|
||||
name: kserve-container
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
requests:
|
||||
cpu: "1"
|
||||
memory: 2Gi
|
||||
protocolVersions:
|
||||
- v1
|
||||
- v2
|
||||
supportedModelFormats:
|
||||
- autoSelect: true
|
||||
name: xgboost
|
||||
priority: 1
|
||||
version: "1"
|
||||
---
|
||||
apiVersion: serving.kserve.io/v1alpha1
|
||||
kind: ClusterStorageContainer
|
||||
metadata:
|
||||
name: default
|
||||
spec:
|
||||
container:
|
||||
image: kserve/storage-initializer:v0.13.0
|
||||
name: storage-initializer
|
||||
resources:
|
||||
limits:
|
||||
cpu: "1"
|
||||
memory: 1Gi
|
||||
requests:
|
||||
cpu: 100m
|
||||
memory: 100Mi
|
||||
supportedUriFormats:
|
||||
- prefix: gs://
|
||||
- prefix: s3://
|
||||
- prefix: hdfs://
|
||||
- prefix: webhdfs://
|
||||
- regex: https://(.+?).blob.core.windows.net/(.+)
|
||||
- regex: https://(.+?).file.core.windows.net/(.+)
|
||||
- regex: https?://(.+)/(.+)
|
||||
File diff suppressed because it is too large
Load Diff
File diff suppressed because one or more lines are too long
@@ -0,0 +1,6 @@
|
||||
apiVersion: kustomize.config.k8s.io/v1beta1
|
||||
kind: Kustomization
|
||||
resources:
|
||||
# Install Kserve in kubeflow namespace
|
||||
- kserve_kubeflow.yaml
|
||||
- kserve-cluster-resources.yaml
|
||||
@@ -0,0 +1 @@
|
||||
ingressGateway=kubeflow/kubeflow-gateway
|
||||
@@ -0,0 +1,44 @@
|
||||
apiVersion: apps/v1
|
||||
kind: Deployment
|
||||
metadata:
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
spec:
|
||||
replicas: 1
|
||||
selector:
|
||||
matchLabels:
|
||||
app.kubernetes.io/component: kserve-models-web-app
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app.kubernetes.io/component: kserve-models-web-app
|
||||
spec:
|
||||
containers:
|
||||
- image: kserve/models-web-app:latest
|
||||
imagePullPolicy: Always
|
||||
name: kserve-models-web-app
|
||||
envFrom:
|
||||
- configMapRef:
|
||||
name: kserve-models-web-app-config
|
||||
ports:
|
||||
- containerPort: 5000
|
||||
name: http
|
||||
livenessProbe:
|
||||
httpGet:
|
||||
path: /healthz/liveness
|
||||
port: http
|
||||
initialDelaySeconds: 0
|
||||
periodSeconds: 10
|
||||
timeoutSeconds: 1
|
||||
failureThreshold: 3
|
||||
successThreshold: 1
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /healthz/readiness
|
||||
port: http
|
||||
initialDelaySeconds: 0
|
||||
periodSeconds: 10
|
||||
timeoutSeconds: 1
|
||||
failureThreshold: 3
|
||||
successThreshold: 1
|
||||
serviceAccountName: kserve-models-web-app
|
||||
@@ -0,0 +1,21 @@
|
||||
apiVersion: networking.istio.io/v1beta1
|
||||
kind: VirtualService
|
||||
metadata:
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
spec:
|
||||
gateways:
|
||||
- knative-serving/knative-ingress-gateway
|
||||
hosts:
|
||||
- '*'
|
||||
http:
|
||||
- match:
|
||||
- uri:
|
||||
prefix: /kserve-endpoints/
|
||||
rewrite:
|
||||
uri: /
|
||||
route:
|
||||
- destination:
|
||||
host: kserve-models-web-app.kserve.svc.cluster.local
|
||||
port:
|
||||
number: 80
|
||||
@@ -0,0 +1,20 @@
|
||||
resources:
|
||||
- rbac.yaml
|
||||
- service.yaml
|
||||
- deployment.yaml
|
||||
- istio.yaml
|
||||
|
||||
namespace: kserve
|
||||
|
||||
commonLabels:
|
||||
kustomize.component: kserve-models-web-app
|
||||
app.kubernetes.io/component: kserve-models-web-app
|
||||
|
||||
images:
|
||||
- name: kserve/models-web-app
|
||||
newName: kserve/models-web-app
|
||||
newTag: v0.13.0-rc.0
|
||||
configMapGenerator:
|
||||
- name: kserve-models-web-app-config
|
||||
literals:
|
||||
- APP_DISABLE_AUTH="True"
|
||||
@@ -0,0 +1,68 @@
|
||||
apiVersion: v1
|
||||
kind: ServiceAccount
|
||||
metadata:
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
---
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRole
|
||||
metadata:
|
||||
name: kserve-models-web-app-cluster-role
|
||||
rules:
|
||||
- apiGroups:
|
||||
- authorization.k8s.io
|
||||
resources:
|
||||
- subjectaccessreviews
|
||||
verbs:
|
||||
- create
|
||||
- apiGroups:
|
||||
- ""
|
||||
resources:
|
||||
- namespaces
|
||||
- pods
|
||||
- pods/log
|
||||
- events
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
- apiGroups:
|
||||
- serving.kserve.io
|
||||
resources:
|
||||
- inferenceservices
|
||||
- inferenceservices/status
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
- watch
|
||||
- create
|
||||
- delete
|
||||
- deletecollection
|
||||
- patch
|
||||
- update
|
||||
- apiGroups:
|
||||
- serving.knative.dev
|
||||
resources:
|
||||
- services
|
||||
- services/status
|
||||
- routes
|
||||
- routes/status
|
||||
- configurations
|
||||
- configurations/status
|
||||
- revisions
|
||||
- revisions/status
|
||||
verbs:
|
||||
- get
|
||||
- list
|
||||
---
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRoleBinding
|
||||
metadata:
|
||||
name: kserve-models-web-app-binding
|
||||
roleRef:
|
||||
apiGroup: rbac.authorization.k8s.io
|
||||
kind: ClusterRole
|
||||
name: kserve-models-web-app-cluster-role
|
||||
subjects:
|
||||
- kind: ServiceAccount
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
@@ -0,0 +1,14 @@
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
spec:
|
||||
ports:
|
||||
- name: http
|
||||
port: 80
|
||||
protocol: TCP
|
||||
targetPort: 5000
|
||||
selector:
|
||||
app.kubernetes.io/component: kserve-models-web-app
|
||||
type: ClusterIP
|
||||
@@ -0,0 +1,42 @@
|
||||
apiVersion: kustomize.config.k8s.io/v1beta1
|
||||
kind: Kustomization
|
||||
|
||||
# Adds namespace to all resources.
|
||||
namespace: kubeflow
|
||||
|
||||
# Labels to add to all resources and selectors.
|
||||
commonLabels:
|
||||
app: kserve
|
||||
app.kubernetes.io/name: kserve
|
||||
|
||||
bases:
|
||||
- ../../base
|
||||
- web-app-authorization-policy.yaml
|
||||
|
||||
patchesStrategicMerge:
|
||||
- patches/web-app-sidecar.yaml
|
||||
|
||||
patchesJson6902:
|
||||
- target:
|
||||
group: networking.istio.io
|
||||
version: v1beta1
|
||||
kind: VirtualService
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
path: patches/web-app-vsvc.yaml
|
||||
|
||||
generatorOptions:
|
||||
disableNameSuffixHash: true
|
||||
|
||||
# To make namespace for standalone installation kustomizable,
|
||||
# variabalize ingress gateway, webhook service name and
|
||||
# kserve namespace in webhook configurations
|
||||
configMapGenerator:
|
||||
- name: kserve-models-web-app-config
|
||||
behavior: replace
|
||||
literals:
|
||||
- USERID_HEADER=kubeflow-userid
|
||||
- APP_PREFIX=/kserve-endpoints
|
||||
|
||||
configurations:
|
||||
- params.yaml
|
||||
@@ -0,0 +1,4 @@
|
||||
commonLabels:
|
||||
- kind: AuthorizationPolicy
|
||||
path: spec/selector/matchLabels
|
||||
create: true
|
||||
+10
@@ -0,0 +1,10 @@
|
||||
apiVersion: apps/v1
|
||||
kind: Deployment
|
||||
metadata:
|
||||
name: kserve-models-web-app
|
||||
namespace: kserve
|
||||
spec:
|
||||
template:
|
||||
metadata:
|
||||
annotations:
|
||||
sidecar.istio.io/inject: "true"
|
||||
+14
@@ -0,0 +1,14 @@
|
||||
- op: replace
|
||||
path: /spec/gateways
|
||||
value:
|
||||
- kubeflow/kubeflow-gateway
|
||||
- op: replace
|
||||
path: /spec/http/0/route/0/destination
|
||||
value:
|
||||
host: kserve-models-web-app.kubeflow.svc.cluster.local
|
||||
port:
|
||||
number: 80
|
||||
- op: replace
|
||||
path: /spec/gateways
|
||||
value:
|
||||
- kubeflow/kubeflow-gateway
|
||||
+18
@@ -0,0 +1,18 @@
|
||||
apiVersion: security.istio.io/v1beta1
|
||||
kind: AuthorizationPolicy
|
||||
metadata:
|
||||
labels:
|
||||
kustomize.component: kserve-models-web-app
|
||||
app.kubernetes.io/component: kserve-models-web-app
|
||||
name: kserve-models-web-app
|
||||
spec:
|
||||
action: ALLOW
|
||||
selector:
|
||||
matchLabels:
|
||||
kustomize.component: kserve-models-web-app
|
||||
app.kubernetes.io/component: kserve-models-web-app
|
||||
rules:
|
||||
- from:
|
||||
- source:
|
||||
principals:
|
||||
- cluster.local/ns/istio-system/sa/istio-ingressgateway-service-account
|
||||
@@ -0,0 +1,6 @@
|
||||
{
|
||||
"instances": [
|
||||
[6.8, 2.8, 4.8, 1.4],
|
||||
[6.0, 3.4, 4.5, 1.6]
|
||||
]
|
||||
}
|
||||
@@ -0,0 +1,4 @@
|
||||
pytest>=7.0.0
|
||||
kserve>=0.12.1
|
||||
kubernetes>=18.20.0
|
||||
requests>=2.18.4
|
||||
@@ -0,0 +1,60 @@
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
import os
|
||||
|
||||
from kubernetes import client
|
||||
from kubernetes.client import V1ResourceRequirements
|
||||
|
||||
from kserve import (
|
||||
constants,
|
||||
KServeClient,
|
||||
V1beta1InferenceService,
|
||||
V1beta1InferenceServiceSpec,
|
||||
V1beta1PredictorSpec,
|
||||
V1beta1SKLearnSpec,
|
||||
)
|
||||
from utils import KSERVE_TEST_NAMESPACE
|
||||
from utils import predict
|
||||
|
||||
|
||||
def test_sklearn_kserve():
|
||||
service_name = "isvc-sklearn"
|
||||
predictor = V1beta1PredictorSpec(
|
||||
min_replicas=1,
|
||||
sklearn=V1beta1SKLearnSpec(
|
||||
storage_uri="gs://kfserving-examples/models/sklearn/1.0/model",
|
||||
resources=V1ResourceRequirements(
|
||||
requests={"cpu": "50m", "memory": "128Mi"},
|
||||
limits={"cpu": "100m", "memory": "256Mi"},
|
||||
),
|
||||
),
|
||||
)
|
||||
|
||||
isvc = V1beta1InferenceService(
|
||||
api_version=constants.KSERVE_V1BETA1,
|
||||
kind=constants.KSERVE_KIND,
|
||||
metadata=client.V1ObjectMeta(
|
||||
name=service_name, namespace=KSERVE_TEST_NAMESPACE
|
||||
),
|
||||
spec=V1beta1InferenceServiceSpec(predictor=predictor),
|
||||
)
|
||||
|
||||
kserve_client = KServeClient(
|
||||
config_file=os.environ.get("KUBECONFIG", "~/.kube/config")
|
||||
)
|
||||
kserve_client.create(isvc)
|
||||
kserve_client.wait_isvc_ready(service_name, namespace=KSERVE_TEST_NAMESPACE)
|
||||
res = predict(service_name, "./data/iris_input.json")
|
||||
assert res["predictions"] == [1, 1]
|
||||
kserve_client.delete(service_name, KSERVE_TEST_NAMESPACE)
|
||||
@@ -0,0 +1,125 @@
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
import json
|
||||
import logging
|
||||
import os
|
||||
import time
|
||||
from urllib.parse import urlparse
|
||||
|
||||
import requests
|
||||
from kubernetes import client
|
||||
|
||||
from kserve import KServeClient
|
||||
from kserve import constants
|
||||
|
||||
logging.basicConfig(level=logging.INFO)
|
||||
|
||||
KSERVE_NAMESPACE = "kserve"
|
||||
KSERVE_TEST_NAMESPACE = "kserve-test"
|
||||
MODEL_CLASS_NAME = "modelClass"
|
||||
|
||||
|
||||
class M2mTokenNotAvailable(Exception):
|
||||
pass
|
||||
|
||||
|
||||
def get_cluster_ip(name="istio-ingressgateway", namespace="istio-system"):
|
||||
api_instance = client.CoreV1Api(client.ApiClient())
|
||||
service = api_instance.read_namespaced_service(name, namespace)
|
||||
if service.status.load_balancer.ingress is None:
|
||||
cluster_ip = service.spec.cluster_ip
|
||||
else:
|
||||
if service.status.load_balancer.ingress[0].hostname:
|
||||
cluster_ip = service.status.load_balancer.ingress[0].hostname
|
||||
else:
|
||||
cluster_ip = service.status.load_balancer.ingress[0].ip
|
||||
return os.environ.get("KSERVE_INGRESS_HOST_PORT", cluster_ip)
|
||||
|
||||
|
||||
def get_m2m_auth_token(env_name="KSERVE_M2M_TOKEN"):
|
||||
try:
|
||||
return os.environ[env_name]
|
||||
except KeyError:
|
||||
raise M2mTokenNotAvailable(env_name)
|
||||
|
||||
|
||||
def predict(
|
||||
service_name,
|
||||
input_json,
|
||||
protocol_version="v1",
|
||||
version=constants.KSERVE_V1BETA1_VERSION,
|
||||
model_name=None,
|
||||
):
|
||||
with open(input_json) as json_file:
|
||||
data = json.load(json_file)
|
||||
|
||||
return predict_str(
|
||||
service_name=service_name,
|
||||
input_json=json.dumps(data),
|
||||
protocol_version=protocol_version,
|
||||
version=version,
|
||||
model_name=model_name,
|
||||
)
|
||||
|
||||
|
||||
def predict_str(
|
||||
service_name,
|
||||
input_json,
|
||||
protocol_version="v1",
|
||||
version=constants.KSERVE_V1BETA1_VERSION,
|
||||
model_name=None,
|
||||
):
|
||||
kfs_client = KServeClient(
|
||||
config_file=os.environ.get("KUBECONFIG", "~/.kube/config")
|
||||
)
|
||||
isvc = kfs_client.get(
|
||||
service_name,
|
||||
namespace=KSERVE_TEST_NAMESPACE,
|
||||
version=version,
|
||||
)
|
||||
# temporary sleep until this is fixed https://github.com/kserve/kserve/issues/604
|
||||
time.sleep(10)
|
||||
cluster_ip = get_cluster_ip()
|
||||
host = urlparse(isvc["status"]["url"]).netloc
|
||||
headers = {
|
||||
"Host": host,
|
||||
"Content-Type": "application/json",
|
||||
}
|
||||
|
||||
try:
|
||||
token = get_m2m_auth_token()
|
||||
headers.update({"Authorization": f"Bearer {token}"})
|
||||
logging.info("M2M Token Found.")
|
||||
except M2mTokenNotAvailable:
|
||||
logging.warn("M2M Token Not found, client authentication disabled.")
|
||||
|
||||
if model_name is None:
|
||||
model_name = service_name
|
||||
|
||||
url = f"http://{cluster_ip}/v1/models/{model_name}:predict"
|
||||
if protocol_version == "v2":
|
||||
url = f"http://{cluster_ip}/v2/models/{model_name}/infer"
|
||||
|
||||
logging.info("Sending Header = %s", headers)
|
||||
logging.info("Sending url = %s", url)
|
||||
logging.info("Sending request data: %s", input_json)
|
||||
response = requests.post(url, input_json, headers=headers)
|
||||
logging.info(
|
||||
"Got response code %s, content %s", response.status_code, response.content
|
||||
)
|
||||
if response.status_code == 200:
|
||||
preds = json.loads(response.content.decode("utf-8"))
|
||||
return preds
|
||||
else:
|
||||
response.raise_for_status()
|
||||
Reference in New Issue
Block a user