MLOpsPython
MLOpsPython copied to clipboard
Error while trying to access Model folder
ot found SourceFolder: /home/vsts/work/r1/a/_model/object_detection:51
I have selected the model and version from the copy dialog in Azure DevOps
$(System.DefaultWorkingDirectory)/_model/object_detection:51
how do I get access to the model files in the MLOps flow?
Can you post the steps you have taken, and a screenshot or copy of the error you are seeing?
yields
expected behavior
-
"copy files" is able to connect to the model folder
-
in case MS devs fail on the first one, just hide the model folder from the available sources.
@j-so is it enough information? Or you just asked
Thank you for the screenshots! We are taking a look.
In the meantime, my recommendation is to use Pipelines instead of Releases (check out the latest version of this repo - getting started). This approach uses the Azure ML CLI, instead of prebuilt release tasks. So it has more flexibility and debuggability.