Closed arpitjasa-db closed 1 year ago
It looks good to me. That's a lot of condition changes.
As one more step of sanity check, could you generate the mlflow recipes project and fs project using both master and your branch? Then do a diff of the two projects? Also do a diff of the training notebooks since they are renamed now.
It looks good to me. That's a lot of condition changes.
As one more step of sanity check, could you generate the mlflow recipes project and fs project using both master and your branch? Then do a diff of the two projects? Also do a diff of the training notebooks since they are renamed now.
Here's Feature Store: https://github.com/databricks/test-delta-stacks/pull/4 and MLflow Recipes: https://github.com/databricks/test-delta-stacks/pull/5 when compared to Delta Tables.
Here's Feature Store: https://github.com/databricks/test-delta-stacks/pull/6 and MLflow Recipes: https://github.com/databricks/test-delta-stacks/pull/7 when compared to the current version pulled from the main branch of Stacks at this time.
I'm having issues with this upgrade, cookiecutter no longer works and the suggested databricks cli version is not available
@mabreuortega The databricks cli version is available here https://github.com/databricks/cli/releases/tag/v0.204.0
@mabreuortega Created PR to explicitly add the link to databricks cli installation doc and releases page https://github.com/databricks/mlops-stack/pull/100
We update MLOps Stacks to use Delta Tables by default instead of MLflow Recipes.
Tested using generated stack here: https://github.com/databricks/test-delta-stacks, and confirmed conflicting options cannot be selected: