ayyoubmaul / ml-regression-flask

ml-regression-flask

Steps by step for running locally

1. Install dependencies using pip install -r requirements.txt
2. Run python3 models/regression.py to run and test the model
3. Run python3 server/regression.py to run endpoint server reggresion

Run with docker

1. docker build -t regression-flask:latest .
2. docker-compose up --build -d

Run in docker container

3. docker exec -it regression-flask bash
4. cd models
5. python3 regression.py

Run in local

3. Export environment variables in terminal session

export AWS_ACCESS_KEY_ID=minio
export AWS_SECRET_ACCESS_KEY=minio123
export MLFLOW_S3_ENDPOINT_URL=http://localhost:9000

5. Open models/regression.py
6. Change mlflow.set_tracking_uri("http://web:5000") to mlflow.set_tracking_uri("http://localhost:5000")
7. Change dataset = pd.read_csv('/app/data/salary.csv') to dataset = pd.read_csv('data/salary.csv')
8. python3 models/regression.py
9. Get run id from MLFlow UI
10. Open browser and go to http://localhost:5002?exp={experience}&run_id={run_id}