0.34.0
This release comes with major upgrades to the python library as well as the dashboard:
- You can now store you secrets in a centralized way instead of having them
tied to a secrets manager stack component. The secrets manager component is deprecated but will still
work while we continue migrating all secrets manager flavors to be available as a backend to store centralized
secrets. Check out the docs
for more information. - Pipelines are now versioned: ZenML detects changes to your steps and structure of your pipelines and
automatically creates new pipeline versions for you. - You can now build the required Docker images for your pipeline without actually running it with
thezenml pipeline build
command. This build can later be used to run the pipeline using the
zenml pipeline run
command or by passing it topipeline.run()
in python. DockerSettings
can now be specified separately for each step of your pipeline. This allows you to build
specialized Docker images for each step in your pipeline if needed.- Metadata for runs and artifacts is now displayed in the dashboard: When viewing a pipeline run in the dashboard,
click on a step or artifact to get useful metadata like the endpoint where your model is deployed or
statistics about your training data.
What's Changed
- Move inbuilt Flavors into the Database by @AlexejPenner in #1187
- Bump
ruff
version to 241 by @strickvl in #1289 - Add docs for run name templates by @schustmi in #1290
- Remove excess help text for
zenml connect
command by @strickvl in #1291 - Increase default service timeout to 60 by @safoinme in #1294
- increase timeout on quickstart example by @safoinme in #1296
- Add warning about MacOS not being supported by @strickvl in #1303
- Always include .zen in docker builds by @schustmi in #1292
- Add warning and docs update for
label_studio
installation issue by @strickvl in #1299 - Loosen version requirements for Great Expectations integration by @strickvl in #1302
- Change zenml init --template to optionally prompt and track email by @stefannica in #1298
- Update docs for Neptune experiment tracker integration by @strickvl in #1307
- Fix the destroy function on the stack recipe CLI by @wjayesh in #1301
- Add missing flavor migrations, make workspace ID optional by @schustmi in #1315
- Bump ruff 246 by @strickvl in #1316
- Remove tag from image name in gcp image builder by @schustmi in #1317
- Fix docs typo by @strickvl in #1318
- Fix step parameter merging by @schustmi in #1320
- Increase timeout for mlflow deployment example by @strickvl in #1308
- Workspace/projects fix for dashboard URL output when running pipeline by @strickvl in #1322
- Component Metadata Tracking Docs by @fa9r in #1319
- Add user environment
zenml info
command to CLI for debugging by @strickvl in #1312 - Added caching to quickstart by @htahir1 in #1321
- Renovation of the zenstore tests by @AlexejPenner in #1275
- Fixes GCP docs typo by @luckri13 in #1327
- Remove deprecated CLI options by @strickvl in #1325
- GCP Image Builder network by @gabrielmbmb in #1323
- improved flavor docs by @htahir1 in #1324
- Commands to register, build and run pipelines from the CLI by @schustmi in #1293
- Validate kserve model name by @strickvl in #1304
- Fix post-execution run sorting by @schustmi in #1332
- Secrets store with SQL back-end by @stefannica in #1313
New Contributors
- @luckri13 made their first contribution in #1327
Full Changelog: 0.33.0...0.34.0