- Install dependencies using
pip install -r requirements.txt
- Run
python3 models/regression.py
to run and test the model - Run
python3 server/regression.py
to run endpoint server reggresion
docker build -t regression-flask:latest .
docker-compose up --build -d
docker exec -it regression-flask bash
cd models
python3 regression.py
- 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
- Open
models/regression.py
- Change
mlflow.set_tracking_uri("http://web:5000")
tomlflow.set_tracking_uri("http://localhost:5000")
- Change
dataset = pd.read_csv('/app/data/salary.csv')
todataset = pd.read_csv('data/salary.csv')
python3 models/regression.py
- Get run id from MLFlow UI
- Open browser and go to
http://localhost:5002?exp={experience}&run_id={run_id}