Week 09 — CI/CD for Machine Learning

Each commit must be potentially deployable. The CI/CD/CT discipline — where the third T is *continuous training*.

MLOps  ·  schedule  ·  Week 09 of 12 ·  ← 08 ·  10 →

Week 09 — CI/CD for Machine Learning

Each commit must be potentially deployable. The CI/CD/CT discipline — where the third T is *continuous training*.

Lecture

Continuous integration for ML (Humble-Farley 2010 applied) · CI for code, data validation, and model performance · GitHub Actions / GitLab CI patterns for ML · model promotion gates · the continuous-training trigger (data drift, schedule, manual).

Read before the lecture

Problem set

Memo 2 — Design a CI/CD pipeline

  1. Design the CI/CD/CT pipeline for the model deployed in Lab 5. Diagram the steps, choose the tools, document the failure recovery. Memo, 1000 words.

Reference text for this week: chapter 09 of the bilingual notes — EN PDF · FR PDF.