v0.11.0

• Batching & caching Feast retrieval from standard transformer
• Enable retry for polling get batch job right after create batch job
• User define function(UDF) for standard transformer (geohash, nested json, s2id)
• Support UDF syntax in UI
• Model version labelling and filtering

v0.11.0-rc2

• Fix caching key when deal with same entity name but different project

v0.11.0-rc1

• Batching & caching Feast retrieval from standard transformer
• Enable retry for polling get batch job right after create batch job
• User define function(UDF) for standard transformer (geohash, nested json, s2id)
• Support UDF syntax in UI
• Model version labelling and filtering

Release v0.10.1

Improvements:

• Standard transformer optimization
• Retry if create prediction job failed

Release v0.10.0

New features:

• Add Request & Response Logging feature
• Add Standard Transformer (Feast Enrichment) feature

Fixes:

• Fixes the pyspark app's dependencies

Release v0.10.0-rc2

Add Standard Transformer (Feast Enrichment) feature.

Release v0.10.0-rc1

Add Request & Response Logging feature and bug fixes on the pyspark app python dependencies

UI Hotfix & Pagination Support

Support API Pagination On Model Version API (#40)

* Support pagination for model versions API

* Update golang client and python client

* Fix code that use merlin api client

* Fixing type check

* Address PR Review:
1. Generalized pagination implementation
2. Remove unnecessary code

* Support pagination on version list page

* Fix warning on ui

* Reformat generated enum named from swagger

* Fix yarn.lock

* Remove unncessary code

* Add comment to exported method OkWithHeaders

* Support search based on environment name

* Load current state after user deploy or serve model

Merlin release v0.9.1

Bump python sdk version (#43)

Merlin release v0.9.0

Merlin sdk 0.9.0.dev0 (#38)

* Publish merlin-sdk 0.9.0.dev0