PyPI · #3071 most downloaded on PyPI

mleap

0.25.21.7M downloads/mocombust/mleap ↗New releases in your daily digest, changelog included.

MLeap Python API

Last release 2 months ago
21 Jul 2026
Release timing varies
gaps range from 2 weeks to 1.1 years
Nearly every release is documented
notes for 22 of 24 stable releases
Nothing withdrawn
no release was ever pulled
9 years old
24 releases · first in 2017

Release timeline

24 releases since 2017

2017201920212023

Releases

0.25.221 Jul 2026
docs: sync and clean up module READMEs by @ltrottier-yelp in #904
What's Changed

docs: sync and clean up module READMEs by @ltrottier-yelp in #904

Upgrade to Spark 4.1.2 and Scala 2.13.18 (MLeap 0.25.2) by @ltrottier-yelp in #905

Full Changelog: v0.25.1...v0.25.2
Open source →
0.25.110 Jul 2026
Upgrade to Spark 4.1.1 and TensorFlow 2.16.2 (MLeap 0.25.1) by @ltrottier-yelp in #903
What's Changed

Upgrade to Spark 4.1.1 and TensorFlow 2.16.2 (MLeap 0.25.1) by @ltrottier-yelp in #903

Full Changelog: v0.25.0...v0.25.1

0.24.010 Mar 2026

Add python 3.7 to 3.12 support by @austinzh in #889

What's Changed

Add python 3.7 to 3.12 support by @austinzh in #889
Upgrade to Scala 2.13.16, Spark 4.0.1, Java 17, and XGBoost 2.0.3 by @ltrottier-yelp in #892
Upgrade to Scala 2.13.16, Spark 4.0.1, Java 17, and XGBoost 2.0.3 by @Urvashi2303 in #887

New Contributors

@Urvashi2303 made their first contribution in #887

Full Changelog: v0.23.3...v0.24.0

Open source →

Additional notes

What's Changed

Add python 3.7 to 3.12 support by @austinzh in https://github.com/combust/mleap/pull/889
Upgrade to Scala 2.13.16, Spark 4.0.1, Java 17, and XGBoost 2.0.3 by @ltrottier-yelp in https://github.com/combust/mleap/pull/892
Upgrade to Scala 2.13.16, Spark 4.0.1, Java 17, and XGBoost 2.0.3 by @Urvashi2303 in https://github.com/combust/mleap/pull/887

New Contributors

@Urvashi2303 made their first contribution in https://github.com/combust/mleap/pull/887

Full Changelog: https://github.com/combust/mleap/compare/v0.23.3...v0.24.0

Open source →

0.23.227 Nov 2024

Fixed an issue with OneHotEncoderOp when computing categorySizes by @ltrottier-yelp in #873

What's Changed

Fixed an issue with OneHotEncoderOp when computing categorySizes by @ltrottier-yelp in #873

New Contributors

@ltrottier-yelp made their first contribution in #873

Full Changelog: v0.23.1...v0.23.2

Open source →

0.23.114 Nov 2023

Add readme for building by @ronandoolan2 in #858

What's Changed

Add readme for building by @ronandoolan2 in #858
Add setClassLoader method to MleapReflection by @austinzh in #816
Ground work for making mleap scala 2.13 compatible by @dotbg in #864
Fix - Add zip slip validation by @David-Fadida in #866
Bump urllib3 from 1.26.5 to 1.26.17 in /python by @dependabot in #868

New Contributors

@ronandoolan2 made their first contribution in #858
@dotbg made their first contribution in #864
@David-Fadida made their first contribution in #866

Full Changelog: v0.23.0...v0.23.1

Open source →

0.23.023 Jun 2023

Upgrade to XGBoost 1.7.3 and serialize objective for XGBoostClassificationModel by @harupy in https://github.com/combust/mleap/pull/847

What's Changed

Upgrade to XGBoost 1.7.3 and serialize objective for XGBoostClassificationModel by @harupy in https://github.com/combust/mleap/pull/847
Mleap Runtime serialization using wrong file name for DecisionTree models by @AllanaEvans in https://github.com/combust/mleap/pull/843
Allow casting a tensor of size 1 to scalar by @praj-0 in https://github.com/combust/mleap/pull/850
Upgrade tensorflow-java to 0.5.0 which corresponds to tensorflow 2.10.1. by @jsleight in https://github.com/combust/mleap/pull/851
Upgrade to Spark 3.4.0. by @jsleight in https://github.com/combust/mleap/pull/852

New Contributors

@harupy made their first contribution in https://github.com/combust/mleap/pull/847
@AllanaEvans made their first contribution in https://github.com/combust/mleap/pull/843

Full Changelog: https://github.com/combust/mleap/compare/v0.22.0...v0.23.0

Open source →

0.22.024 Feb 2023

Allow only one of input or default in MathBinary by @praj-0 in https://github.com/combust/mleap/pull/841

What's Changed

Allow only one of input or default in MathBinary by @praj-0 in https://github.com/combust/mleap/pull/841
Upgrade python to 3.7/3.8 and scikit-learn to v1.0.0 by @jsleight in https://github.com/combust/mleap/pull/836
Upgrade spark dependency to 3.3.0 by @WeichenXu123 in https://github.com/combust/mleap/pull/824

New Contributors

@praj-0 made their first contribution in https://github.com/combust/mleap/pull/841

Full Changelog: https://github.com/combust/mleap/compare/v0.21.0...v0.22.0

Open source →

Additional notes

Newer versions are tracked via github's release notes.

Open source →

0.21.122 Dec 2022

This release backports a fix to Java 11 classpaths for spark 3.2 users.

What's Changed

This release backports a fix to Java 11 classpaths for spark 3.2 users.

Alter jvm_scala_object to better handle nested scala classes. by @jsleight in https://github.com/combust/mleap/pull/837

Full Changelog: https://github.com/combust/mleap/compare/v0.21.0...v0.21.1

Open source →

0.21.02 Nov 2022

Shade log4j for databricks fat jar by @WeichenXu123 in https://github.com/combust/mleap/pull/812

What's Changed

Shade log4j for databricks fat jar by @WeichenXu123 in https://github.com/combust/mleap/pull/812
ScalaPB & sbt Upgrade by @emitc2h in https://github.com/combust/mleap/pull/818
Upgrade xgboost dependency to be 1.6.1 version by @WeichenXu123 in https://github.com/combust/mleap/pull/822
Additional Math unary operations by @shyamsunder00 in https://github.com/combust/mleap/pull/826
Added backward compatibility fix for XGBoost models by @chaitanya-basava in https://github.com/combust/mleap/pull/829
Address MathBinary issue with zeros defaults by @arthurarj in https://github.com/combust/mleap/pull/828

New Contributors

@emitc2h made their first contribution in https://github.com/combust/mleap/pull/818
@shyamsunder00 made their first contribution in https://github.com/combust/mleap/pull/826
@chaitanya-basava made their first contribution in https://github.com/combust/mleap/pull/829
@arthurarj made their first contribution in https://github.com/combust/mleap/pull/828

Full Changelog: https://github.com/combust/mleap/compare/v0.20.0...v0.21.0

Open source →

Additional notes

Breaking Changes

MLeap is now built with Java 11 instead of Java 8.

What's Changed

Shade log4j for databricks fat jar by @WeichenXu123 in https://github.com/combust/mleap/pull/812
ScalaPB & sbt Upgrade by @emitc2h in https://github.com/combust/mleap/pull/818
Upgrade xgboost dependency to be 1.6.1 version by @WeichenXu123 in https://github.com/combust/mleap/pull/822
Additional Math unary operations by @shyamsunder00 in https://github.com/combust/mleap/pull/826
Added backward compatibility fix for XGBoost models by @chaitanya-basava in https://github.com/combust/mleap/pull/829
Address MathBinary issue with zeros defaults by @arthurarj in https://github.com/combust/mleap/pull/828

Open source →

0.20.04 Apr 2022

Fixes 'cannot assign instance of java.lang.invoke.SerializedLambda' for Tensorflow transformer when using sparkTransform() method

Bug Fixes

Fixes 'cannot assign instance of java.lang.invoke.SerializedLambda' for Tensorflow transformer when using sparkTransform() method (#801)

Improvements

Upgrade to xgboost v1.5.2 (#799,#803)
Upgrade to spark v3.2.0 (#799)
Upgrade to tensorflow java 0.4.0 and tensorflow 2.7.1 (#802)
Add undoLog cleanup around usages of scala.reflect.api.Types.TypeApi.<:< (#806)

Open source →

0.19.012 Jan 2022

MapType is now supported as a core data type

New Features

MapType is now supported as a core data type (#789)
MapEntrySelector is a new Transformer that can be used to select values from maps (#789)
Spark Vectors can now be cast into MLeap Tensors (#791)

Bug Fixes

Fixes uid parity issues for certain Spark 3 transformers (#788)
MLeap Tensor -> Spark Vector casting logic is fixed for non-increasing indices (#794)

Improvements

Upgrades to xgboost v1.3.1 (#778)
Updates shading rules for databricks runtime assembly (#780)
StringIndexerModel now performs faster lookups by caching the index size (#793)
Upgrades springboot version to 2.6.2 and junit to 5.8.2

Open source →

0.18.013 Sept 2021

Scala 2.11 is no longer supported due to Spark 3 upgrade

Breaking Changes

Scala 2.11 is no longer supported due to Spark 3 upgrade
Please use 0.18.1 mleap version if you have models serialized with both Spark 2.x and Spark 3.x

New Features

MathBinaryModel now supports Logit operations
TensorflowTransformerOp now supports serialization and deserialization using SavedModel format
Casting.cast now supports conversions between ListType and TensorType

Bug Fixes

Fix OneHotEncoder Python serialization

Improvements

Upgrade to scikit-learn 0.22
Upgrade to Spark 3.0.2
Upgrade Tensorflow version to 2.4.1

Open source →

0.17.07 May 2021

upgrade to xgboost 1.0.0 - using h2oai Predictor

Breaking Changes

New Features

upgrade to xgboost 1.0.0 - using h2oai Predictor
support for using xgboost predictor when using xgboost regressor
MathBinaryModel now supports Min and Max operations

Bug Fixes

fix Spark deserialization of random forest classifier to include numTrees

Improvements

scoring optimizations for Interacting and CountVectorizer

Open source →

0.16.022 May 2020

Fix default ports when running grpc/http requests; default grpc port is 65328 and can be overridden via MLEAP_GRPC_PORT; default http port should be:

Breaking Changes

Fix default ports when running grpc/http requests; default grpc port is 65328 and can be overridden via MLEAP_GRPC_PORT; default http port should be: 65327 and can be overridden via MLEAP_HTTP_PORT

New Features

Upgrade to Spark version 2.4.5
Support for a performant implementation of the XGboost runtime (XGboost Predictor)
Scikit-learn support for MultinomialLogisticRegression
Support for min/max values other than defaults (i.e. 0.0 and 1.0) in MinMaxScalerModel
Support for custom transformers (StringMap, MathUnary, MathBinary) in Pyspark
Support MLWritable/MLReadable for custom transformers (StringMap, MathUnary, MathBinary) and fix this for Imputer transformer
Fixes support for loading/storing bundles from/to hdfs in Pyspark
Improve importing mleap version for python modules

Bug Fixes

Fix XGBoost sparse vector support
Fix MinMaxScalerModel outputs different in Spark vs MLeap
Fix Spark deserialization for CountVectorizer transformer
Added support for HandleInvalid in Bucketizer, VectorIndexer
Fix setting HandleInvalid by default to OneHotEncoder for backwards compatibility
Fixes MLReader for Imputer mleap implementation of Spark transformer

Improvements

Minor documentation updates

Open source →

0.15.025 Nov 2019

Load models at start up in mleap-spring-boot

Breaking Changes

None

New Features

Load models at start up in mleap-spring-boot
Add support for Python 3
StringMap transformer - add new optional parameters handleInvalid & defaultValue
Add support for LinearSVC transformer/model

Bug Fixes

Fix Tensorflow bundle writing when transform() method isn't necessarily called
Fix FrameReader reading a very large mleap frame

Improvements

Update xgboost4j and fix databricks runtime
Use openjdk:8-jre-slim as docker base image
Bump urllib3 from 1.23 to 1.24.2 in python package
Add default grpc port to docker config
General documentation improvements

Open source →