Feast Versions Save

The Open Source Feature Store for Machine Learning

v0.25.2

1 year ago

0.25.2 (2022-10-06)

Bug Fixes

  • Fix docker image for feature-server (#3272) (4ce366a)
  • Fix Feast UI release process to update the feast-ui package (#3267) (d118fe4)
  • Stream feature view meta undefined created_timestamp issue (#3266) (efbd4b0)
  • Udf in stream feature view UI shows pickled data (#3268) (f4a83a7)
  • Updated quickstart notebook to patch an incorrect reference to an outdated featureview name (#3271) (7fa4bbf)

v0.25.1

1 year ago

0.25.1 (2022-09-30)

Bug Fixes

  • Add X-Trino-Extra-Credential header and remove user override (#3246) (f38506c)
  • Add postgres to the feature server Dockerfile to fix helm chart flow (#3261) (a2cb995)
  • Add stream feature view in the Web UI (#3257) (47d4c93)
  • Build dockerfile correctly (#3239) (a2dc0d0)
  • Configuration to stop coercion of tz for entity_df (#3255) (fdc8d67)
  • Enable users to upgrade a batch source into a push source (#3213) (8f2fb58)
  • Return 422 on bad push source name (#3214) (8abbcd9)
  • Stream feature view not shown in the UI (#3251) (55e28e2)
  • Update snowflake materialization messages (#3230) (a63d440)
  • Use configured user in env var instead of "user" for Trino (#3254) (e7ed3d5)

v0.25.0

1 year ago

0.25.0 (2022-09-20)

This release adds bug fixes and some new functionality:

  • Spark materialization engine.
  • to_remote_storage capability for Spark offline store.

Bug Fixes

  • Broken Feature Service Link (#3227) (e117082)
  • Feature-server image is missing mysql dependency for mysql registry (#3223) (ae37b20)
  • Fix handling of TTL in Go server (#3232) (f020630)
  • Fix materialization when running on Spark cluster. (#3166) (175fd25)
  • Fix push API to respect feature view's already inferred entity types (#3172) (7c50ab5)
  • Fix release workflow (#3144) (20a9dd9)
  • Fix Shopify timestamp bug and add warnings to help with debugging entity registration (#3191) (de75971)
  • Handle complex Spark data types in SparkSource (#3154) (5ddb83b)
  • Local staging location provision (#3195) (cdf0faf)
  • Remove bad snowflake offline store method (#3204) (dfdd0ca)
  • Remove opening file object when validating S3 parquet source (#3217) (a906018)
  • Snowflake config file search error (#3193) (189afb9)
  • Update Snowflake Online docs (#3206) (7bc1dff)

Features

  • Add to_remote_storage functionality to SparkOfflineStore (#3175) (2107ce2)
  • Add ability to give boto extra args for registry config (#3219) (fbc6a2c)
  • Add health endpoint to py server (#3202) (43222f2)
  • Add snowflake support for date & number with scale (#3148) (50e8755)
  • Add tag kwarg to set Snowflake online store table path (#3176) (39aeea3)
  • Add workgroup to athena offline store config (#3139) (a752211)
  • Implement spark materialization engine (#3184) (a59c33a)

v0.24.1

1 year ago

0.24.1 (2022-09-08)

Bug Fixes

  • Fix push API to respect feature view's already inferred entity types (#3172) (5dba4a3)
  • Fix release workflow (#3144) (c77a655)
  • Fix Shopify timestamp bug and add warnings to help with debugging entity registration (#3191) (95066ea)
  • Handle complex Spark data types in SparkSource (#3154) (0e13573)

v0.24.0

1 year ago

0.24.0 (2022-08-25)

This release adds many bug fixes as well as new functionality:

  • Snowflake and Bytewax materialization engines.
  • Azure provider with Synapse offline store and Azure registry blob store (moved into contrib from an external repo).
  • A Cassandra online store.
  • Athena offline store.

Bug Fixes

  • Check if on_demand_feature_views is an empty list rather than None for snowflake provider (#3046) (9b05e65)
  • FeatureStore.apply applies BatchFeatureView correctly (#3098) (41be511)
  • Fix Feast Java inconsistency with int64 serialization vs python (#3031) (4bba787)
  • Fix feature service inference logic (#3089) (4310ed7)
  • Fix field mapping logic during feature inference (#3067) (cdfa761)
  • Fix incorrect on demand feature view diffing and improve Java tests (#3074) (0702310)
  • Fix Java helm charts to work with refactored logic. Fix FTS image (#3105) (2b493e0)
  • Fix on demand feature view output in feast plan + Web UI crash (#3057) (bfae6ac)
  • Fix release workflow to release 0.24.0 (#3138) (a69aaae)
  • Fix Spark offline store type conversion to arrow (#3071) (b26566d)
  • Fixing Web UI, which fails for the SQL registry (#3028) (64603b6)
  • Force Snowflake Session to Timezone UTC (#3083) (9f221e6)
  • Make infer dummy entity join key idempotent (#3115) (1f5b1e0)
  • More explicit error messages (#2708) (e4d7afd)
  • Parse inline data sources (#3036) (c7ba370)
  • Prevent overwriting existing file during persist (#3088) (69af21f)
  • Register BatchFeatureView in feature repos correctly (#3092) (b8e39ea)
  • Return an empty infra object from sql registry when it doesn't exist (#3022) (8ba87d1)
  • Teardown tables for Snowflake Materialization testing (#3106) (0a0c974)
  • UI error when saved dataset is present in registry. (#3124) (83cf753)
  • Update sql.py (#3096) (2646a86)
  • Updated snowflake template (#3130) (f0594e1)

Features

  • Add authentication option for snowflake connector (#3039) (74c75f1)
  • Add Cassandra/AstraDB online store contribution (#2873) (feb6cb8)
  • Add Snowflake materialization engine (#2948) (f3b522b)
  • Adding saved dataset capabilities for Postgres (#3070) (d3253c3)
  • Allow passing repo config path via flag (#3077) (0d2d951)
  • Contrib azure provider with synapse/mssql offline store and Azure registry store (#3072) (9f7e557)
  • Custom Docker image for Bytewax batch materialization (#3099) (cdd1b07)
  • Feast AWS Athena offline store (again) (#3044) (989ce08)
  • Implement spark offline store offline_write_batch method (#3076) (5b0cc87)
  • Initial Bytewax materialization engine (#2974) (55c61f9)
  • Refactor feature server helm charts to allow passing feature_store.yaml in environment variables (#3113) (85ee789)

v0.23.2

1 year ago

0.23.2 (2022-08-15)

Bug Fixes

  • Fix field mapping logic during feature inference (#3067) (eb885b1)
  • Fix incorrect on demand feature view diffing and improve Java tests (#3074) (0ff0ec4)
  • Fix on demand feature view output in feast plan + Web UI crash (#3057) (a32d247)
  • Fix Spark offline store type conversion to arrow (#3071) (a49f70c)

v0.22.4

1 year ago

0.22.4 (2022-08-15)

Bug Fixes

  • Fix field mapping logic during feature inference (#3067) (3668702)
  • Fix incorrect on demand feature view diffing and improve Java tests (#3074) (dd46d45)
  • Fix on demand feature view output in feast plan + Web UI crash (#3057) (a44fe66)
  • Fix Spark offline store type conversion to arrow (#3071) (8e6a6b1)

v0.22.3

1 year ago

0.22.3 (2022-08-10)

Bug Fixes

  • Check if on_demand_feature_views is an empty list rather than None for snowflake provider (#3046) (67af727)
  • Fixing Web UI, which fails for the SQL registry (#3028) (56d645c)
  • More explicit error messages (#2708) (93c1c15)
  • Move gcp back to 1.47.0 since grpcio-tools 1.48.0 got yanked from pypi (#2990) (f7e44da)
  • Return an empty infra object from sql registry when it doesn't exist (#3022) (329bc47)

v0.23.1

1 year ago

0.23.1 (2022-08-10)

Bug Fixes

  • Check if on_demand_feature_views is an empty list rather than None for snowflake provider (#3046) (6af401f)
  • Fix Feast Java inconsistency with int64 serialization vs python (#3031) (f340aeb)
  • Fixing Web UI, which fails for the SQL registry (#3028) (d584ecd)
  • More explicit error messages (#2708) (ca48963)
  • Return an empty infra object from sql registry when it doesn't exist (#3022) (9a64e77)

v0.23.0

1 year ago

0.23.0 (2022-08-02)

This release adds many bug fixes as well as new functionality:

  • Support for a Snowflake online store
  • A batch materialization engine interface, which now makes materialization a pluggable component. There are two reference implementations already (local, which was the former default, and an AWS Lambda engine which scales up materialization with Lambda)
  • A to_remote_storage method in RetrievalJob (the output of get_historical_features). This makes methods that were already implemented (e.g. to_gcs or to_s3) now visible at the interface level. The Lambda materialization engine implementations use these methods to unload data to a bucket before scalably loading this data into the online store.
  • Improvements to Feast Web UI + SavedDatasets

Features

  • Add an experimental lambda-based materialization engine (#2923) (6f79069)
  • Add column reordering to write_to_offline_store (#2876) (8abc2ef)
  • Add custom JSON table tab w/ formatting (#2851) (0159f38)
  • Add CustomSourceOptions to SavedDatasetStorage (#2958) (23c09c8)
  • Add Go option to feast serve command (#2966) (a36a695)
  • Add interfaces for batch materialization engine (#2901) (38b28ca)
  • Add pages for individual Features to the Feast UI (#2850) (9b97fca)
  • Add snowflake online store (#2902) (f758f9e), closes #2903
  • Add Snowflake online store (again) (#2922) (2ef71fc), closes #2903
  • Add to_remote_storage method to RetrievalJob (#2916) (109ee9c)
  • Support retrieval from multiple feature views with different join keys (#2835) (056cfa1)

Bug Fixes

  • Add dummy alias to pull_all_from_table_or_query (#2956) (5e45228)
  • Bump version of Guava to mitigate cve (#2896) (51df8be)
  • Change numpy version on setup.py and upgrade it to resolve dependabot warning (#2887) (80ea7a9)
  • Change the feature store plan method to public modifier (#2904) (0ec7d1a)
  • Deprecate 3.7 wheels and fix verification workflow (#2934) (040c910)
  • Do not allow same column to be reused in data sources (#2965) (661c053)
  • Fix build wheels workflow to install apache-arrow correctly (#2932) (bdeb4ae)
  • Fix file offline store logic for feature views without ttl (#2971) (26f6b69)
  • Fix grpc and update protobuf (#2894) (86e9efd)
  • Fix night ci syntax error and update readme (#2935) (b917540)
  • Fix nightly ci again (#2939) (1603c9e)
  • Fix the go build and use CgoArrowAllocator to prevent incorrect garbage collection (#2919) (130746e)
  • Fix typo in CONTRIBUTING.md (#2955) (8534f69)
  • Fixing broken links to feast documentation on java readme and contribution (#2892) (d044588)
  • Fixing Spark min / max entity df event timestamps range return order (#2735) (ac55ce2)
  • Move gcp back to 1.47.0 since grpcio-tools 1.48.0 got yanked from pypi (#2990) (fc447eb)
  • Refactor testing and sort out unit and integration tests (#2975) (2680f7b)
  • Remove hard-coded integration test setup for AWS & GCP (#2970) (e4507ac)
  • Resolve small typo in README file (#2930) (16ae902)
  • Revert "feat: Add snowflake online store (#2902)" (#2909) (38fd001)
  • Snowflake_online_read fix (#2988) (651ce34)
  • Spark source support table with pattern "db.table" (#2606) (3ce5139), closes #2605
  • Switch mysql log string to use regex (#2976) (5edf4b0)
  • Update gopy to point to fork to resolve github annotation errors. (#2940) (ba2dcf1)
  • Version entity serialization mechanism and fix issue with int64 vals (#2944) (d0d27a3)