MLOps with Azure ML

159751

update arm template to make workspace sku configurable (#283)

MLOps with Azure ML

158098

Simplify docs flow (#297)

MLOps with Azure ML

Move instruction to install AML extension to Azure Devops setup instr…

…uctions (#298)

MLOps with Azure ML

153255

update azureml sdk (#287)

MLOps with Azure ML

149223

Replaced Env class with dataclass (#277)

MLOps with Azure ML

Add Terraform option to environment_setup (#268)

* setup basic folder and file structure

* add tf backend file and bash script to create state storage

* basic pipeline for infrastructure with tf - yaml, tf, bash

* naming and deleting unnecessary bash script

* updated documentation

* added to the get_started.md guide

* added terraform plan step

MLOps with Azure ML

Update SDK to 1.3.0 (#266)

Fixes #265.

MLOps with Azure ML

Fix docker pipeline by removing trailing whitespace (#264)

The docker pipeline fails to tag because the trailing whitespace gets included in the tag name.

MLOps with Azure ML

Update CI conda deps to match training/scoring SDK (#263)

- Tied SDK version to 1.2.x as with conda_dependencies.yml
- Lock versions to point updates
- Kept the rest of the deps manually specified to keep image size small and minimize regressions

3.1.0 Release

In 3.1.0, we have several new features to enable more customization in the DevOps pipeline. We also have cleaned up the naming of our pipelines and restructured our docs for a better onboarding experience.

Features:

• Enable deploying models registered by previous builds (skip first two stages of pipeline) #207 @jotaylo
• Improve environment customization process #206 @algattik
• Add reusable AzureML Environments #217 @sudivate
• Enable versioned datasets #218 @eedorenko
• Allow users to specify model tags in parameters.json #237 @eedorenko
• Add image tags for pipeline build ID, github release ID, and AzureML SDK version #240 @sudivate
• Set the training step to allow reuse the results from previous runs #140 @sudivate
• Use Model Package for image creation #260 @sudivate
• Run unit tests in any case during pipeline run #199 @sbaidachni
• Clean up pipeline variables and add comments #211 @jotaylo
• Rename pipeline YAML files to a new convention #212 @tcare
• Remove BuildId as a parameter to ML pipeline #214 @jotaylo
• New standalone train.py for training logic outside of AzureML and AzureML logic moved to train_aml.py #219 @jotaylo
• Rename config.json to parameters.json #223 @jotaylo
• Add get_latest_model method to model helper util code #231 @starlord-daniel
• Upgrade AzureML SDK in build agent #235 @eedorenko

Fixes:

• Add code integration guide #243 @jotaylo
• Improve and refactor docs #197 @eedorenko, #216, #221, #220, #225 @tcare, #247 @jotaylo
• Fix input features in model training/scoring #198 @romanlytvyn
• Add basic comment descriptions to YAML pipelines #200 @tcare
• Doc: clarifications in Getting Started #201 @stevehaigh
• Use correct parameters in WebApp deploy #202 @stevehaigh
• Remove 60 minute timeout in ML training pipeline #203 @algattik
• Use pipeline variable instead of hard coded subscription reference #205 @jotaylo
• Fix pipeline condition to trigger ML pipeline #209 @jotaylo
• Fix abtest pipeline #224 @tcare
• Fix environment not updating on change #230 @celaus
• Bootstrap fixes #250 @tcare
• Fix load sample data #252 @cindyweng, #254 @omartin2010
• Fix training/scoring conda dependencie #262 @tcare
• Add abtest YAML pipeline to bootstrap guide #233 @celaus