Lilly Tech Systems
Learn DVC
Master Data Version Control — the open-source tool that brings Git-like versioning to data, models, and ML pipelines for reproducible machine learning.
Your Learning Path
Follow these lessons in order, or jump to any topic that interests you.
1. Introduction
What is DVC, why Git alone isn't enough for ML, and how DVC solves data versioning.
2. Setup & Configuration
Install DVC, initialize a project, configure remote storage (S3, GCS, Azure).
3. Data Versioning
Track data with dvc add, push/pull to remotes, switch between data versions with Git.
4. Pipelines
Define reproducible ML pipelines with dvc.yaml, manage dependencies, and run stages.
5. Experiments
Track experiments, compare metrics, manage parameters, and version model outputs.
6. Best Practices
CI/CD integration, team workflows, storage optimization, and production deployment.
What You'll Learn
By the end of this course, you'll be able to:
Version Data Like Code
Use Git-like commands to version datasets and models, with storage on S3, GCS, or Azure.
Build Pipelines
Create reproducible ML pipelines that automatically track dependencies and outputs.
Run Experiments
Track, compare, and manage ML experiments with metrics and parameter versioning.
Automate with CI/CD
Integrate DVC into GitHub Actions and other CI/CD systems for automated ML workflows.