FLAML Versions Save

Highlights

• Support new OpenAI gpt-3.5-turbo and gpt-4 models in autogen. Thanks to @gagb @kevin666aa @qingyun-wu @ekzhu @BeibinLi .
• MathChat implemented with autogen.agents. Thanks to @kevin666aa @qingyun-wu.
• Time-series related functionality in automl is factored out. Thanks to @EgorKraevTransferwise .

Thanks to all the contributors and reviewers @thinkall @qingyun-wu @EgorKraevTransferwise @kevin666aa @liususan091219 @skzhang1 @jtongxin @pcdeadeasy @markharley @int-chaos !

What's Changed

• Fix documentation by @sonichi in https://github.com/microsoft/FLAML/pull/1075
• encode timeout msg in bytes by @sonichi in https://github.com/microsoft/FLAML/pull/1078
• Add pands requirement in benchmark option by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1070
• Fix pyspark tests in workflow by @thinkall in https://github.com/microsoft/FLAML/pull/1071
• Docmentation for agents by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1057
• Links to papers by @sonichi in https://github.com/microsoft/FLAML/pull/1084
• update openai model support by @sonichi in https://github.com/microsoft/FLAML/pull/1082
• string to array by @sonichi in https://github.com/microsoft/FLAML/pull/1086
• Factor out time series-related functionality into a time series Task object by @EgorKraevTransferwise in https://github.com/microsoft/FLAML/pull/989
• An agent implementation of MathChat by @kevin666aa in https://github.com/microsoft/FLAML/pull/1090

New Contributors

• @kevin666aa made their first contribution in https://github.com/microsoft/FLAML/pull/1090

Full Changelog: https://github.com/microsoft/FLAML/compare/2.0.0rc1...v2.0.0rc2

This release includes:

• A Major Refactor: the creation of an automl option to remove unnecessary dependencies for autogen and tune (thanks to @sonichi.)
• A newly added blog post addressing adaptation in HumanEval (thanks to @sonichi.)
• A newly added tutorials folder containing all the tutorials on FLAML (thanks to @qingyun-wu, @sonichi, and @thinkall.)
• Documentation Improvement and Link Corrections.
• The addition of documentation and a notebook example on interactive LLM agents in FLAML (thanks to @qingyun-wu, @sonichi, @thinkall, and @pcdeadeasy.)
• Support more azure openai api_type (thanks to @thinkall )
• Suppress warning message of pandas_on_spark to_spark (thanks to @thinkall )
• Support shell command and multiple code blocks (thanks to @sonichi )
• Improve the system message for assistant agent (thanks to @sonichi and @gagb )
• Improve utility functions for config lists (thanks to @sonichi )
• Reuse docker image in a session (thanks to @sonichi and @gagb )

A hearty welcome to our new contributor, @badjouras, who made their first contribution. Thanks to code reviewers @gagb @pcdeadeasy @liususan091219 @thinkall @levscaut @sonichi @qingyun-wu.

What's Changed

• Blogpost for adaptation in HumanEval by @sonichi in https://github.com/microsoft/FLAML/pull/1048
• Improve messaging in documentation by @sonichi in https://github.com/microsoft/FLAML/pull/1050
• create an automl option to remove unnecessary dependency for autogen and tune by @sonichi in https://github.com/microsoft/FLAML/pull/1007
• docs: 📝 Fix link to installation section in Task-Oriented-AutoML.md by @badjouras in https://github.com/microsoft/FLAML/pull/1051
• doc and test update by @sonichi in https://github.com/microsoft/FLAML/pull/1053
• remove redundant doc and add tutorial by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1004
• add agent notebook and documentation by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1052
• Support more azure openai api_type by @thinkall in https://github.com/microsoft/FLAML/pull/1059
• suppress warning message of pandas_on_spark to_spark by @thinkall in https://github.com/microsoft/FLAML/pull/1058
• Agent notebook example with human feedback; Support shell command and multiple code blocks; Improve the system message for assistant agent; Improve utility functions for config lists; reuse docker image by @sonichi in https://github.com/microsoft/FLAML/pull/1056

New Contributors

• @badjouras made their first contribution in https://github.com/microsoft/FLAML/pull/1051

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.2.4...2.0.0rc1

This release contains:

• improved support for using a list of configurations (thanks to @BeibinLi ),
• using a filter to select from the responses out of a sequence of configurations (doc).
• a new experimental human-proxy agent (thanks to @qingyun-wu and @gagb).
• utility function to create config lists.
• method clear_cache added in oai.Completion.
• update of the default search space (thanks to @Kyoshiin and @LittleLittleCloud ).
• prepare for flaml v2 (thanks to @qingyun-wu for writing the blogpost).

Breaking change:

• cache_path is renamed into cache_path_root in set_cache.

Thanks to code reviewers @skzhang1 @jtongxin @pcdeadeasy @ZviBaratz @LittleLittleCloud @Borda , and to @liususan091219 @thinkall for fixing test error.

What's Changed

• Catch AuthenticationError trying different configs by @BeibinLi in https://github.com/microsoft/FLAML/pull/1023
• chat completion check by @sonichi in https://github.com/microsoft/FLAML/pull/1024
• update model of text summarization in test by @liususan091219 in https://github.com/microsoft/FLAML/pull/1030
• Human agent by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1025
• fix of website link by @sonichi in https://github.com/microsoft/FLAML/pull/1042
• Blogpost by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1026
• Update default search space by @Kyoshiin in https://github.com/microsoft/FLAML/pull/1044
• Fix PULL_REQUEST_TEMPLATE and improve test by removing unnecessary environment variable by @thinkall in https://github.com/microsoft/FLAML/pull/1043
• response filter by @sonichi in https://github.com/microsoft/FLAML/pull/1039

New Contributors

• @BeibinLi made their first contribution in https://github.com/microsoft/FLAML/pull/1023
• @Kyoshiin made their first contribution in https://github.com/microsoft/FLAML/pull/1044

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.2.3...v1.2.4

This release contains a number of updates in autogen and automl.

• We added more utilities and documentation improvement to flaml.oai such as logging, templating and using multiple configs to make developing and experimentation with OpenAI models more convenient. (thanks to @afourney @victordibia @torronen @ekzhu )
• We added an experimental coding agent based on GPT-4. (thanks to @BeibinLi @qingyun-wu @skzhang1 )
• We added options to disable the default mlflow logging in automl. (thanks to @garar)
• We make better use of parallelism in auto-scale spark cluster for automl and tune. (thanks to @thinkall )

Thanks to @royninja @luckyklyist @Borda @qingyun-wu @thinkall @liususan091219 for other bug fixes, documentation improvement, and engineering improvement. Thanks to @victordibia @skzhang1 @kevin666aa @jtongxin @levscaut for code reviews.

What's Changed

• version update post release v1.2.2 by @sonichi in https://github.com/microsoft/FLAML/pull/1005
• fixing the typo #990 by @royninja in https://github.com/microsoft/FLAML/pull/994
• fixed sentence misplace #998 by @luckyklyist in https://github.com/microsoft/FLAML/pull/1010
• pyproject.toml & switch to Ruff by @Borda in https://github.com/microsoft/FLAML/pull/976
• update readme by @qingyun-wu in https://github.com/microsoft/FLAML/pull/1014
• raise content_filter error by @sonichi in https://github.com/microsoft/FLAML/pull/1018
• Fix catboost failure in mac-os python<3.9 by @thinkall in https://github.com/microsoft/FLAML/pull/1020
• coding agent; logging by @sonichi in https://github.com/microsoft/FLAML/pull/1011
• Add mlflow_logging param by @garar in https://github.com/microsoft/FLAML/pull/1015
• fix NLP zero division error by @liususan091219 in https://github.com/microsoft/FLAML/pull/1009
• update max_spark_parallelism to fit in auto-scale spark cluster by @thinkall in https://github.com/microsoft/FLAML/pull/1008
• update spark session in spark tests by @thinkall in https://github.com/microsoft/FLAML/pull/1006
• Mark experimental classes; doc; multi-config trial by @sonichi in https://github.com/microsoft/FLAML/pull/1021

New Contributors

• @luckyklyist made their first contribution in https://github.com/microsoft/FLAML/pull/1010
• @garar made their first contribution in https://github.com/microsoft/FLAML/pull/1015

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.2.2...v1.2.3

What's Changed

• update nlp notebook by @liususan091219 in https://github.com/microsoft/FLAML/pull/940
• Blog post for LLM tuning by @sonichi in https://github.com/microsoft/FLAML/pull/986
• fix zerodivision by @liususan091219 in https://github.com/microsoft/FLAML/pull/1000
• extract code from text; solve_problem; request_timeout in config; improve code by @sonichi in https://github.com/microsoft/FLAML/pull/999

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.2.1...v1.2.2

What's Changed

• handling nlp divide by zero by @liususan091219 in https://github.com/microsoft/FLAML/pull/926
• set black with 120 line length by @Borda in https://github.com/microsoft/FLAML/pull/975
• Post release update by @sonichi in https://github.com/microsoft/FLAML/pull/985
• fix tests failure caused by version incompatibility by @thinkall in https://github.com/microsoft/FLAML/pull/995
• Expose feature and label transformer in automl.py by @JaneIllario in https://github.com/microsoft/FLAML/pull/993
• make context optional; improve error handling and doc by @sonichi in https://github.com/microsoft/FLAML/pull/997

New Contributors

• @JaneIllario made their first contribution in https://github.com/microsoft/FLAML/pull/993

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.2.0...v1.2.1

This is a major release with many new features.

Highlights

• autogen subpackage in preview: text generation optimization, including support for ChatGPT and GPT-4.
• Spark dataframe and SynapseML support in preview.
• Holt-Winters exponential smoothing for time series forecasting.
• Task class abstraction for easier task-specific development.

What's Changed

• precommit: end-of-file-fixer by @Borda in https://github.com/microsoft/FLAML/pull/929
• adding precommit check by @Borda in https://github.com/microsoft/FLAML/pull/930
• Bump http-cache-semantics from 4.1.0 to 4.1.1 in /website by @dependabot in https://github.com/microsoft/FLAML/pull/913
• adding openai documentation by @liususan091219 in https://github.com/microsoft/FLAML/pull/938
• improve max_valid_n and doc by @sonichi in https://github.com/microsoft/FLAML/pull/933
• Update doc and notebook by @skzhang1 in https://github.com/microsoft/FLAML/pull/946
• ChatGPT support by @sonichi in https://github.com/microsoft/FLAML/pull/942
• Extract task class from automl by @markharley in https://github.com/microsoft/FLAML/pull/857
• fix delete in TransformersEstimator, so to remove the ckpt not found warning in PR 940 by @liususan091219 in https://github.com/microsoft/FLAML/pull/950
• Support spark dataframe as input dataset and spark models as estimators by @thinkall in https://github.com/microsoft/FLAML/pull/934
• gpt-4 support; openai workflow fix; model str; timeout; voting by @sonichi in https://github.com/microsoft/FLAML/pull/958
• more tolerant time limit for test_overtime by @levscaut in https://github.com/microsoft/FLAML/pull/960
• Adding a test function for OpenAI completion in flaml by @qingyun-wu in https://github.com/microsoft/FLAML/pull/951
• Noctus contribute by @AyeshGK in https://github.com/microsoft/FLAML/pull/961
• add timeout to openai create by @thinkall in https://github.com/microsoft/FLAML/pull/973
• ci: concurrency by @Borda in https://github.com/microsoft/FLAML/pull/931
• Bump eta from 1.12.3 to 2.0.0 in /website by @dependabot in https://github.com/microsoft/FLAML/pull/941
• Bump webpack from 5.74.0 to 5.76.1 in /website by @dependabot in https://github.com/microsoft/FLAML/pull/952
• Bump @sideway/formula from 3.0.0 to 3.0.1 in /website by @dependabot in https://github.com/microsoft/FLAML/pull/963
• Add Holt-Winters exponential smoothing by @coffepowered in https://github.com/microsoft/FLAML/pull/962
• autogen subpackage by @sonichi in https://github.com/microsoft/FLAML/pull/968

New Contributors

• @AyeshGK made their first contribution in https://github.com/microsoft/FLAML/pull/961

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.1.3...v1.2.0

This version contains multiple updates, including support of Azure OpenAI APIs, user-defined meta features in zero-shot AutoML, and time monitoring for spark jobs. Thanks to all the contributors!

What's Changed

• workflow path->paths by @sonichi in https://github.com/microsoft/FLAML/pull/915
• azure oai by @sonichi in https://github.com/microsoft/FLAML/pull/920
• Added extras for synapse by @thinkall in https://github.com/microsoft/FLAML/pull/916
• Update hf version by @liususan091219 in https://github.com/microsoft/FLAML/pull/918
• bump version to 1.1.3; workflow trigger by @sonichi in https://github.com/microsoft/FLAML/pull/924
• added a documentation about small time budget by @Animaholic in https://github.com/microsoft/FLAML/pull/914
• Fix Dependabot alerts by @skzhang1 in https://github.com/microsoft/FLAML/pull/922
• pre-commit: update config by @Borda in https://github.com/microsoft/FLAML/pull/925
• Improve annotations in automl and ml modules by @coffepowered in https://github.com/microsoft/FLAML/pull/919
• Updated dict type args default value to None by @thinkall in https://github.com/microsoft/FLAML/pull/927
• fix #493 by @LittleLittleCloud in https://github.com/microsoft/FLAML/pull/932
• add PySparkOvertimeMonitor to avoid exceeding time budget by @levscaut in https://github.com/microsoft/FLAML/pull/923
• minor change to notebook by @sonichi in https://github.com/microsoft/FLAML/pull/928

New Contributors

• @Borda made their first contribution in https://github.com/microsoft/FLAML/pull/925
• @LittleLittleCloud made their first contribution in https://github.com/microsoft/FLAML/pull/932
• @levscaut made their first contribution in https://github.com/microsoft/FLAML/pull/923

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.1.2...v1.1.3

News

• AAAI-23 tutorial: We will be giving a tutorial about FLAML in AAAI-23 on Feb 08, 2023. Please find the tutorial agenda in this page.
• Latest research to appear at ICLR-23: Our latest research on multiobjective HPO is accepted as one of the notable-top-5% research papers in ICLR 2023:
  • Please find more details about the method in our paper: Targeted Hyperparameter Optimization with Lexicographic Preferences Over Multiple Objectives. Shaokun Zhang, Feiran Jia, Chi Wang, Qingyun Wu. ICLR 2023 (notable-top-5%).
  • Please find detailed documentation about this new functionality in this page: https://microsoft.github.io/FLAML/docs/Use-Cases/Tune-User-Defined-Function#lexicographic-objectives.
• OpenAI support is added in this release. A notebook example can be found at: https://github.com/microsoft/FLAML/blob/main/notebook/integrate_openai.ipynb

What's Changed

• handle num_samples=-1 by @sonichi in https://github.com/microsoft/FLAML/pull/879
• Bump json5 from 2.2.1 to 2.2.3 in /website by @dependabot in https://github.com/microsoft/FLAML/pull/877
• fix the doc error of customized metrics in automl by @skzhang1 in https://github.com/microsoft/FLAML/pull/883
• Bump ua-parser-js from 0.7.32 to 0.7.33 in /website by @dependabot in https://github.com/microsoft/FLAML/pull/900
• display data head in notebook; exclude None by @sonichi in https://github.com/microsoft/FLAML/pull/885
• Support percentage tolerance for lexicographic optimization by @skzhang1 in https://github.com/microsoft/FLAML/pull/875
• Document how to use the group k-fold by @coffepowered in https://github.com/microsoft/FLAML/pull/894
• update doc for research papers by @qingyun-wu in https://github.com/microsoft/FLAML/pull/912
• stratified group kfold splitter by @sonichi in https://github.com/microsoft/FLAML/pull/899
• add cost budget; move loc of make_dir by @sonichi in https://github.com/microsoft/FLAML/pull/888
• Add bibtex entries for research publications by @Animaholic in https://github.com/microsoft/FLAML/pull/904
• Openai by @sonichi in https://github.com/microsoft/FLAML/pull/905

New Contributors

• @coffepowered made their first contribution in https://github.com/microsoft/FLAML/pull/894
• @Animaholic made their first contribution in https://github.com/microsoft/FLAML/pull/904

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.1.1...v1.1.2

What's Changed

• create dir for log file name by @sonichi in https://github.com/microsoft/FLAML/pull/867
• Do not persist entire AutoMLState in Searcher by @Yard1 in https://github.com/microsoft/FLAML/pull/870
• fix #871: call check_spark only when necessary by @thinkall in https://github.com/microsoft/FLAML/pull/872
• notebook test; spark warning message; reproducibility bug; sequential tuning stop condition by @sonichi in https://github.com/microsoft/FLAML/pull/869

Full Changelog: https://github.com/microsoft/FLAML/compare/v1.1.0...v1.1.1