search

Azure / mlops-v2

Azure MLOps (v2) solution accelerators. Enterprise ready templates to deploy your machine learning models on the Azure Platform.
https://learn.microsoft.com/en-us/azure/machine-learning/concept-model-management-and-deployment
MIT License
514 stars 258 forks source link

[QuickStart.md] Path structure in deploy-infra pipeline is different from current path structure #32

Open yvonnebarthp opened 2 years ago

yvonnebarthp commented 2 years ago

The Pipeline Path can not be recreated since the folder structure in the given repo is different from the screenshot shown in the quickstart. Screenshot from the quickstart: image

Current path structure: image

setuc commented 2 years ago

Hi, We have checked this issue and tried to reproduce. The only potential issue that we can see if you directly work with the repo, for example cloning it, that won't work. You need to use sparse checkout to properly adjust the paths and the folders. See this step in the script sparse_checkout.sh in the root folder of the MLOps-v2 Repository

https://github.com/Azure/mlops-v2/blob/3742042f1dac499060bf389bc603f34058b7762f/sparse_checkout.sh#L25-L27

krisbock commented 2 years ago

hI @setuc , the confusion comes from the wording and the screenshots associated with those steps.
1) The text "Select your /MLOps-Test repository. ("Empty" reposity you created in 2.2)" refers to the template generated repo (referred to as "mono").
2) The screenshots show the "mono" repo that was generated from the template. It should be showing the "sparse" repo created from the script.

This will cause users to select the wrong repo to point ADO to for the initial infra YAML. It also means users will update the wrong config-infra-xxxx.yml.

Suggest sticking to a set of example repo names and having consistent references and images throughout .