project folder for ML model deployment using Kubernetes
This is repo for project for demonstration of deployment of ML model (containerized) using Kubernetes(k8s)
The Docker image is uploaded on https://hub.docker.com . The image can be pulled using
$ docker pull subbu0319/placement-app
The details of folder/files in the project folder is as follows :
(a) Notebooks folder - jupyter noteboook for data cleaning , model building with model and Dictvectorizer(encoding of cat columns) as output
(b) Data - contains train and test data .
(c) predict.py - python script to prodice an API for model prediction using flask framework
(d) predict-test.py - python test script with test features for testing the model prediction . It uses requests libary to access REST API. Please keep in mind to enter correct URL in the script for testing the application (depends on loca/rremote testing)
(e) Pipfile and Pipfile.lock - pipenv generated dependencies files
(f) Dockerfile - for building Docker image
(g) kubeconfig folder - contains deployment.yaml and service.yaml for deploying the container on k8s