This repository contains demo code to showcase the complete pipeline for using MLOps
This Project is based on the cookiecutter data science project template. #cookiecutterdatascience
The demo code requires the following to work:
- AWS access key and secret key stored in the
.aws/credentials
file under the default namespace - The AWS user should have access to Elastic Beanstalk, S3 and EC2 instance running MLFlow in it
- The raw dataset can be found in the UCI website
- Upload the dataset to S3 and configure the bucket name in the
params.yaml
file
- In the AWS configure a EC2 machine with port
8080
open and installmlflow
and instantiate the server - Update the
params.yaml
with the URL to the EC2 machine this will help to set the tracking uri for MLFlow
- After updating the
params.yaml
file with appropriate paths run thedvc repro
command from the project root - The
result app
will be deployed to your AWS Elastic Beanstalk environment