Working on various ML models and need to strategise all metrics for model maturity. Anyone who can share thoughts would be great start?
So far I have researched about following
Number of models available
Number / Type of templates available
Total number of models /pipelines
Code Coverage
BDH scan results
Coverity Defects
Container Vulnerabilities
% Buld Green
%Deploy Success
Deployment Time
Production Lead Time
MTTR
Change Frequency
Error Budget
Cross validationa accuracy
Feature importance
Training Loss
Model Size
Container Health
Precision / Recall
F1 Score
Accuracy
ROC-AUC