Skip to content

Latest commit

 

History

History
101 lines (75 loc) · 3.4 KB

README.md

File metadata and controls

101 lines (75 loc) · 3.4 KB
Raw

Sentiment-KFP

Sentiment KLP is an open source analysis software used to perform sentiment analysis with Kubeflow Pipelines.

Running the vanilla KFP version

  1. Go to Kubeflow Dashboard and on the left panel click on Volumes.

  2. Click on the “+ New Volume” button on the top right and create a volume by giving the name “model-volume”.

  3. Add a PodDefault

kubectl apply -f https://raw.githubusercontent.com/GigliOneiric/Sentiment-KFP/main/config/PodDefault.yml

  1. Go to Kubeflow Dashboard and on the left panel click on Notebooks.

  2. Click on the “+ New Notebook” button on the top right and create a notebook by giving it a name.

  3. Add “Allow access to Kubeflow Pipelines” in Configurations

  4. After the set up is done, click on the Connect button next to the notebook you just created. It will take you to a JupyterLab.

  5. In the JupyterLab launcher start a new terminal session to clone the github repo. In the terminal enter the following commands:

$ git clone https://github.com/GigliOneiric/Sentiment-KFP.git
  1. After succesfully cloning the repo open the notebook named "BiGRU.ipynb" in the folder "models" by double-clicking on this name in the left hand directory structure, and to run it click on the "restart the whole kernel and re-reun the whole notebook" button in the top menu of the notebook.

Optional: Setup Seldon Core

1 . Label the namespace for inference tasks in kubernetes

kubectl label namespace kubeflow-user-example-com serving.kubeflow.org/inferenceservice=enabled
  1. Install s2i
# Create a directory
mkdir /tmp/s2i/ && cd /tmp/s2i/

# Download S2I
curl -s https://api.github.com/repos/openshift/source-to-image/releases/latest| grep browser_download_url | grep linux-amd64 | cut -d '"' -f 4  | wget -qi -

# Unpack the tar
tar xvf source-to-image*.gz

# Move it to /usr/local/bin path to be executable
sudo mv s2i /usr/local/bin
rm -rf /tmp/s2i/
  1. Install Docker
apt install docker.io

  1. Copy the container folder to any directory on kubernetes

  2. Creation of an image within the container directory copied previously

s2i build . seldonio/seldon-core-s2i-python38:1.10.0 seldon-sentiment:latest
  1. Deploying the image in the Docker Registery
# Start the registry
docker run -d -p 5000:5000 --restart=always --name registry registry:2

# Tag the image so that it points to the registry
docker tag seldon-sentiment:latest  localhost:5000/seldon-sentiment:latest 

# Push the image
docker push localhost:5000/seldon-sentiment:latest
  1. Creation of a SeldonDeployment
  • Create the SeldonDeployment
kubectl apply -f https://raw.githubusercontent.com/GigliOneiric/Sentiment-KFP/main/seldon/SeldonDeployment.yml
  • Verify the SeldonDeployment
kubectl get sdep -n kubeflow-user-example-com seldon-sentiment -o json | jq .status
  1. Set up port forwarding
tmux new -d -s seldon "kubectl port-forward $(kubectl get pods -l istio=ingressgateway -n istio-system -o jsonpath='{.items[0].metadata.name}') -n istio-system 8050:8080 --address 0.0.0.0"
  1. Test the SeldonDeployment
  • Read the session cookie from the Kubeflow dashboard in the developer mode of the browser
document.cookie
  • Send a request
curl http://localhost:8050/seldon/kubeflow-user-example-com/seldon-sentiment /api/v0.1/predictions --data-urlencode 'json={"data":{"ndarray":[["This is a test"]]}}' -H "Cookie: authservice_session=YOURCOOKIE" -v