How rcatalanvargas/tarea-1-cloud-computing works
This repository is a FastAPI service for predicting whether a student is placed in employment. The main workflow accepts and validates a student record (or batch), arranges features using metadata, invokes the serialized model, and returns class probabilities. The README also documents model exploration, evaluation, health and model-info endpoints, and interactive API documentation. The training workflow is described but its implementation was not sampled, so its wiring to the deployed artifact is not shown.
Main entry points: main.py Prediction routes · model.pkl Inference pipeline · metadata.json Model metadata
Components 10
Inference API
- Prediction routes
app/main.py - Input contracts
app/main.py - Response builder
app/main.py - Health endpoint
app/main.py - Model info endpoint
app/main.py - Interactive docs
app/main.py
Model artifacts
- Inference pipeline
model/model.pkl - Model metadata
model/metadata.json
Model development
- Data exploration
notebooks/00_Exploracion_inicial.ipynb
Other components
- API client
Connections 14
- API client → Prediction routes: submits prediction
- Prediction routes → Input contracts: validates input
- Prediction routes → Model metadata: selects features
- Prediction routes → Inference pipeline: predicts classes
- Prediction routes → Inference pipeline: estimates probabilities
- Prediction routes → Response builder: builds response
- Response builder → API client: returns results
- API client → Health endpoint: checks status
- API client → Model info endpoint: requests model details
- API client → Interactive docs: opens documentation
- Health endpoint → Inference pipeline: checks loaded model
- Model info endpoint → Model metadata: reads details
- Prediction routes → Inference pipeline: loads at startup
- Prediction routes → Model metadata: loads at startup