Challenge | Automation solution | Keep It Simple Solution |
---|---|---|
Training-Serving Skew, where differences crop up between training and serving stacks | Feature Store | Transform Pattern |
Clients are interested in a capability that involves multiple steps, only one of which is the ML model | ML Pipelines | Stateless Serving Function |
The incoming data changes characteristics | Continuous evaluation | Scheduled evaluation |
The correct answer changes over time | Continuous training | Systematic triggers |
Model fairness issues arise | Mumbo Jumbo | Manual sliced evaluations and other checks |
Parts of the system aren’t changed when new code is checked in | CI/CD | From-scratch builds |
Need to be able to troubleshoot old results | Model registry, Data registry | Version control all deployed artifacts |
Last active
September 18, 2022 13:38
-
-
Save lakshmanok/ce6ec434293649d6313c4a549c3761d9 to your computer and use it in GitHub Desktop.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment