Build a basic mL pipeline using open source technology today
Tech used in this pipeline:
- Python
- Docker/Kubernetes
- Mlflow
- Minio
- Seldon Core
- Github
- Jenkins
- ArgoCD
- Go into experiments
- Run
docker-compose up -d
to start up the underlying services
Training the Experiment with mlFlow
- Run
docker exec -it mlflow_server python train.py
to kick off the training and experiments
Build API Endpoint for model prediction
Test with microservice:
seldon-core-microservice MyModel REST --service-type MODEL