Releases: getindata/kedro-vertexai
Releases · getindata/kedro-vertexai
Release 0.11.1
[0.11.1] - 2024-07-01
�
Release 0.11.0
What's Changed
- Applied copier template config for consistency among plugins (mostly cicd config)
- Updated dependencies and tested for kedro 0.19.3
- Node grouping: Changed convention from : to . due to introduced kedro limitation on special characters in node tags
- Removed EnvTemplatedConfigLoader that gets replaced by default OmegaConf capabilities
Full Changelog: 0.10.0...0.11.0
Release 0.10.0
[0.10.0] - 2023-11-22
- Added explicite pyarrow dependency to avoid critical vulnerability
- Updated dependencies and tested for kedro
0.18.14
- [Feature 🚀] Node grouping: added option to group multiple Kedro nodes together at execution in single Vertex AI process to allow better optimization - less steps, shorter delays while running Vertex AI nodes and less wasted time of data serialization thanks to possibility to use the MemoryDataset
�
Release 0.9.1
[0.9.1] - 2023-08-16
- Updated dependencies of kedro to
0.18.8
, mlflow to2.3.2
and others - Upgrade dependencies to resolve GHSA-6628-q6j9-w8vg.
�
Release 0.9.0
[0.9.0] - 2023-05-15
- Add cache to Kedro's context in the
ContextHelper
class to prevent re-loading - Upgrade dependencies to support
kedro>=0.18.8
- Add support for
OmegaConfigLoader
- Upgrade misc. dependencies
- Remove deprecated
KedoVertexAIConfigLoaderHook
⚠️ Change default behaviour of config loader in the plugin to rely on project's one instead ofEnvTemplatedConfigLoader
- [Docs 📝] Update documentation
- Improve E2E tests config
�
Release 0.8.1
[0.8.1] - 2022-12-30
- Add cache to Kedro's context in the
ContextHelper
class to prevent re-loading
�
Release 0.8.0
[0.8.0] - 2022-12-09
- Added support for configuration of resources and node selectors with Kedro node tags
- Added support for gpu configuration on Vertex AI (by adding
node_selectors
section andgpu
resources entry invertexai.yml
configuration file) - Added --auto-build option to run-once that calls 'docker build' and 'docker push' for you before running the job on VertexAI. It introduces '--yes' option to disable confirmation prompt
�
Release 0.7.0
[0.7.0] - 2022-09-08
- Add better MLflow authorization entrypoints (via Hooks and MLflow Request Header Provider Plugin)
�
Release 0.6.0
[0.6.0] - 2022-08-22
- Add auto-dataset creation, to make intermediate dataset creation transparent to the end-user (no need to explicitly add them in the Data Catalog) (#8)
�