12 Factors of Reproducible Machine Learning in Production
A set of guiding principles to help you better productionize your machine learning models.
Sep 28, 202010 Mins Read
A set of guiding principles to help you better productionize your machine learning models.
![Why ML in production is (still) broken - [#MLOps2020]](https://assets.zenml.io/webflow/64a817a2e7e2208272d1ce30/79950cfa/65316d2bc051413294285f1e_mlopsworldthumbnail.png)
The MLOps movement and associated new tooling is starting to help tackle the very real technical debt problems associated with machine learning in production.
Pipelines help you think and act better when it comes to how you execute your machine learning training workflows.
Using config files to specify infrastructure for training isn't widely practiced in the machine learning community, but it helps a lot with reproducibility.
ZenML makes it easy to setup training pipelines that give you all the benefits of cached steps.
Software engineering best practices have not been brought into the machine learning space, with the side-effect that there is a great deal of technical debt in these code bases.
