How datatalksclub/machine-learning-zoomcamp works
Machine Learning Zoomcamp is a hands-on course whose central learner workflow runs from framing a problem through preparing data, training and evaluating models, then exposing and deploying them as services. The supplied tree emphasizes deployment examples, including model training, HTTP prediction, Kubernetes/TensorFlow Serving, serverless functions, and KServe. The graph shows those documented course stages and the source-visible calls; it does not imply that separate lesson examples form one executable pipeline. Several deployment implementations are indexed but not excerpted, so their internal wiring is left unspecified.
Main entry points: README.md Course materials · code/train.py Model training · app.py Prediction API · 05-deployment/train.py Lead model training
Components 16
Course workflow
- Course materials
README.md - Problem framing
README.md
Model development
- Data preparation
05-deployment/code/train.py - Model training
05-deployment/code/train.py - Model evaluation
05-deployment/code/train.py - Lead features
cohorts/2026/homework/05-deployment/model.py - Lead model training
cohorts/2026/homework/05-deployment/train.py - Model artifacts
cohorts/2026/homework/05-deployment/train.py
Prediction services
- Prediction API
10-kubernetes/workshop/service/app.py - Prediction gateway
10-kubernetes/code/gateway.py - Tensor encoding
10-kubernetes/code/proto.py - Model service
05-deployment/code/predict.py
Deployment platforms
- Lambda inference
09-serverless/code/lambda_function.py - KServe examples
11-kserve/code - Production deployment
README.md
Other components
- Learner
Connections 19
- Learner → Course materials: follows
README.md - Course materials → Problem framing: teaches framing
README.md - Course materials → Data preparation: teaches preparation
README.md - Course materials → Model training: teaches training
README.md - Course materials → Model evaluation: teaches evaluation
README.md - Course materials → Prediction API: teaches API serving
README.md - Course materials → Production deployment: teaches deployment
README.md - Data preparation → Model training: prepares inputs
05-deployment/code/train.py - Model training → Model evaluation: measures AUC
05-deployment/code/train.py - Model training → Lead features: normalizes records
- Lead model training → Lead features: uses normalization
cohorts/2026/homework/05-deployment/train.py - Lead model training → Model artifacts: writes artifacts
cohorts/2026/homework/05-deployment/train.py - Learner → Prediction API: submits image URL
10-kubernetes/workshop/service/app.py - Prediction API → Production deployment: serves predictions
10-kubernetes/workshop/service/app.py - Learner → Prediction gateway: sends prediction request
10-kubernetes/code/gateway.py - Prediction gateway → Tensor encoding: encodes tensor
10-kubernetes/code/gateway.py - Course materials → Lambda inference: covers Lambda
README.md - Course materials → KServe examples: covers KServe
- Course materials → Model service: covers prediction services
README.md