MLOps
- The process on turning raw data into an actual machine learning model
- MLOps applies to the entire lifecycle - from integrating with model generation (software development lifecycle, continuous integration/continuous delivery), orchestration, and deployment, to health, diagnostics, governance, and business metrics
ML Lifecycle
| Order | MLOps Steps | DevOps Steps |
|---|---|---|
| 1 | Train | Code |
| 2 | Save | Build |
| 3 | Test | Test |
| 4 | Serve | Deploy |
| 5 | Monitor | Monitor |
- Save
-
Onnx
-
Test
- Canary Rollout
VertexAI
- MLops as a service
- Handle this process as a service
- Available on Google Cloud
- Steps
- Ingest
- Analyze
- Transform
- Train
- Model
- Evaluate
- Deploy
- Predict
Hugging Face
- MLops as a service
- Offers also pre-trained open source models to be used as starting point