Experimenting with training a model in different ways within two cloud service providers
- push training image to GAR
- create vertex training job
- view results?
EXP 1: sklearn dummy
Model: sample_sklearn_model Compute Service: Vertex AI Compute hardware: n1-standard-4 (w/ 100gb ssd) Provisioning time: 1 minute 30 seconds Task time: 30 seconds Total time: 2 minutes Cost: Notes: startup time was rather long. Not good for quick exp.
**EXP 2: