Replies: 10 comments 9 replies
-
Possible transient error when creating the dataset, solution is to wait a few minutes and retry |
Beta Was this translation helpful? Give feedback.
-
Use existing SP/RG/Workspace. For this, an example JSON is needed ,see below: { |
Beta Was this translation helpful? Give feedback.
-
Local CLI auth - need to set correct subscription and login |
Beta Was this translation helpful? Give feedback.
-
Clarify dataset vs datastore for training data |
Beta Was this translation helpful? Give feedback.
-
VSCode is not necessarily needed, a development environment is all that's really needed. We might want to mention this in the setup/pre work. |
Beta Was this translation helpful? Give feedback.
-
The SP creation/secret content is repeated in challenge 2 |
Beta Was this translation helpful? Give feedback.
-
Wording in challenge 4 is confusing. There's no "pipeline" created in the previous steps, just triggering a job, yet challenge 4 asks them to run the "pipeline" |
Beta Was this translation helpful? Give feedback.
-
Have a part 2 OpenLab that focuses on model monitoring, data drift, etc |
Beta Was this translation helpful? Give feedback.
-
Having AML API coverage as a pre-req would be helpful |
Beta Was this translation helpful? Give feedback.
-
Add a data scientist focused aspect |
Beta Was this translation helpful? Give feedback.
-
Suggestions:
Replace portal ui resource group create commands to cli:
https://microsoftlearning.github.io/mslearn-mlops/documentation/00-set-up.html
az group create --location <location_region> --name rg-dev-mlops
Beta Was this translation helpful? Give feedback.
All reactions