Machine Learning Models Versioning Using Open Source Tools
summary
Shows a concrete workflow for tracking models and artifacts using open source pieces: git lfs for small model files, CI pipelines that capture reproducible artifacts, and an artifact-storage alternative when models hit tens of gigabytes. Explains repository layout, commitable metadata, and how to reproduce a model from a specific commit. Demonstrates tradeoffs—what breaks (git lfs at scale), what does work (artifact pipelines + automated checks), and practical fixes used in production.
AI-written summary · about
chapters (9)
- 0:00ML model versioning intro
- 3:02Managing ML model files
- 8:01Git LFS for models
- 12:27Data management best practices
- 17:21Artifact tracking & automation
- 21:48Optimizing versioning workflows
- 25:31Open source ML versioning tools
- 28:03Case studies & transformations
- 31:59Key takeaways
related talks
24mModel Versioning: Why, When, and How
32mReproducibility and versioning of ML systems
01h 25mBest practices and solution design using data versioning for machine learning
37mOpen Source Tools for ML Experiments Management
59mDevOps for Machine Learning Projects
31mGetting ML Right in a Complex Data World
57mManaging Machine Learning Lifecycle with MLflow
38mML Data Version Control and Reproducibility at Scale