0.7.5

Small bugfixes and testing additions.

What's Changed

• Add link to Github repository in TOC by @strickvl in #91
• Add CodeQL checks by @strickvl in #90
• Add unit tests by @strickvl in #80
• Add missing Vertex flavor by @strickvl in #94
• Fix "Permission Denied" error while executing dig script by @wjayesh in #93
• Fix reference to empty eks resource when creating S3 by @wjayesh in #92
• Stop CI from running on push to develop branch by @strickvl in #95
• Release 0.7.5 by @strickvl in #96

Full Changelog: 0.7.4...0.7.5

0.7.4

Small bug fixes. Some documentation updates as well, linked to how mlstacks works together with ZenML.

What's Changed

• Update variable name for release publication token by @strickvl in #85
• Renaming the environment variable for analytics by @bcdurak in #87
• Update zenml-mlstacks sample command by @strickvl in #89
• Handle errors more gracefully by @strickvl in #88
• Use real Segment analytics package instead of deprecated version by @strickvl in #86

Full Changelog: 0.7.3...0.7.4

0.7.3

What's Changed

• Add zenml <-> mlstacks compatibility check to CI by @strickvl in #84
• Add missing aws flavor (bugfix) by @strickvl in #83

Full Changelog: 0.7.2...0.7.3

0.7.2

Package dependency fixes for zenml compatibility.

What's Changed

• Bump incompatible packages by @strickvl in #82

Full Changelog: 0.7.1...0.7.2

0.7.1

This is a minor release that fixes the Python base requirement and relaxes our PyYaml dependency (to allow you to use mlstacks in the same environment as zenml).

What's Changed

• Add Python requirement for package by @strickvl in #79
• Bump pyyaml for ZenML by @strickvl in #81

Full Changelog: 0.7.0...0.7.1

0.7.0

Lots has changed since our last release! (Our name even changed!)

Stacks are now deployed by way of a Python package.

Full information is available at our new documentation site https://mlstacks.zenml.io.

What's Changed

• mlstacks package and associated logic by @strickvl in #67

Full Changelog: 0.6.1...0.7.0

0.6.1

What's Changed

• fix k3d recipe var names by @wjayesh in #57
• update vpc version for AWS stacks by @MarkusSagen in #59
• Fixing the links after the major docs rehab by @bcdurak in #62
• Fix typos on CI by @strickvl in #63
• Create a "zenml" ns with K8S orch by @wjayesh in #61
• Replace ACLs with Bucket policies for S3 by @wjayesh in #60
• Fix host names for AWS recipes by @wjayesh in #69
• Fix S3 creation and access from MLflow by @wjayesh in #68

New Contributors

• @MarkusSagen made their first contribution in #59
• @bcdurak made their first contribution in #62

Full Changelog: 0.6.0...0.6.1

0.6.0

🧙 ZenML Stack Component Deploy now live!

This release brings important additions to the modular recipes, including a rewrite of Kubernetes cluster modules across all modular recipes, addition of more components like Vertex and Sagemaker to the list of available services, along with a ton of bug fixes.

What's Changed

• [BUGFIX] Replace length check with null check for kubectl by @wjayesh in #39
• Transition important local values to vars by @wjayesh in #40
• Update name of Vertex recipe to gcp-vertexai from vertex-ai by @fa9r in #42
• [BUGFIX] Fix mlflow reference when disabled in Vertex recipe by @wjayesh in #43
• add gcloud auth requirement to prerequisites by @wjayesh in #46
• Fix Kubernetes provider config (check null and fix dependency) by @wjayesh in #48
• Add spellcheck to CI by @strickvl in #47
• Test branch protection rule by @wjayesh in #49
• Revert "Test branch protection rule" by @wjayesh in #50
• Add sagemaker as step operator by @wjayesh in #51
• Add info about role binding to airflow readme by @wjayesh in #52
• sync default region across gcp and aws by @wjayesh in #53
• Randomize output names for stack components by @wjayesh in #54
• Add modular VertexAI step operator and orchestrator as part of gcp-modular by @strickvl in #55
• Fix typo by @strickvl in #56

New Contributors

• @fa9r made their first contribution in #42

Full Changelog: 0.5.0...0.6.0

0.4.0

Welcome Modular Recipes 🥳🎉

This release introduces the concept of modular recipes, starting with AWS, GCP and local K3D deployments. Modular recipes allow the user to selectively enable different components of their stack from the same recipe, eliminating the need to manually extend existing recipes for achieving a different combination than the already provided recipes.

Going forward, modular recipes will completely replace the other recipes as the user will be able to achieve the same output using one or the other combination of stack components. This is also a major step forward in achieving greater ease of deployment of individual stack components, from the CLI and dashboard both.

What's Changed

• run tests on develop by @wjayesh in #27
• Add GCP Airflow recipe using Cloud Composer by @wjayesh in #26
• Modularized recipes by @stefannica in #30
• make enable a variable instead of local by @wjayesh in #32
• K3D minimal recipe by @safoinme in #28

New Contributors

• @stefannica made their first contribution in #30
• @safoinme made their first contribution in #28

Full Changelog: 0.3.1...0.4.0