Add a manifest json to 3.7 so that the featurizer in ml workbench can read in the path of included algorithms at runtime. This allows the algo repo to update independently without worrying about breaking the featurizer.
PR Checklist
[x] Added comments for more comprehensible code.
[NA] Added unit and/or regression tests.
[ ] (Optional) Notified Doc/Prod team, if this PR contains any new features.
PR Summary
PR Checklist
What is the behavior change?
Other information:
Related PRs: