mlctl notebook experience

[p1] As as data scientist I run mlctl init, point to specific cells in my notebook as a training entry point, and be able to deploy a training job from within the cell.
[p2] As a data scientist, when I run a training job from my notebook, I can retrieve the model from my notebook for analysis in the notebook
[p2] As a data scientist, when I invoke a hosting deployment from my notebook and model I can access in my notebook environment.