v2.0.0

Prepare for a roller coaster ride of innovation with the launch of FLAML v2.0.0! This is not just another update but a culmination of numerous enhancements, novel features, and exciting improvements we've made from v2.0.0rc1 to v2.0.0rc5, leading to the grand v2.0.0 release.

• With v2.0.0rc1, we embarked on a major refactor with the creation of an [automl] option to declutter dependencies for autogen and tune.
• In v2.0.0rc2, we supercharged FLAML with support for new OpenAI gpt-3.5-turbo and gpt-4 models in autogen and rolled out the extensibility of autogen agents.
• With v2.0.0rc3, we upped the ante by adding new OpenAI models' support of functions in agents and provided a handy code example in a dedicated notebook.
• v2.0.0rc4 brought a host of improvements to the agentchat framework, enabling many new applications.
• v2.0.0rc5 pushed the boundaries further by making auto-reply methods pluggable and supporting an asynchronous mode in agents.

Finally, we arrive at the grand v2.0.0 release! This version boasts of numerous feature enhancements in autogen, like multi-agent chat framework (in preview), expanded OpenAI model support, enhanced integration with Spark, and much more.

Documentation for AutoGen: https://microsoft.github.io/FLAML/docs/Use-Cases/Autogen
Examples: https://microsoft.github.io/FLAML/docs/Examples/AutoGen-AgentChat
Blogposts: https://microsoft.github.io/FLAML/blog

A huge shoutout to @qingyun-wu @kevin666aa @skzhang1 @ekzhu @BeibinLi @thinkall @LittleLittleCloud @JieyuZ2 @gagb @EgorKraevTransferwise @markharley @int-chaos @levscaut @feiran-jia @liususan091219 @royninja @pcdeadeasy as well as our new contributors @badjouras, @LeoLjl, @xiaoboxia, and @minghao51 who joined us during this journey. Your contributions have played a pivotal role in shaping this release.

What's Changed

• Blogpost for adaptation in HumanEval by @sonichi in #1048
• Improve messaging in documentation by @sonichi in #1050
• create an automl option to remove unnecessary dependency for autogen and tune by @sonichi in #1007
• docs: 📝 Fix link to installation section in Task-Oriented-AutoML.md by @badjouras in #1051
• doc and test update by @sonichi in #1053
• remove redundant doc and add tutorial by @qingyun-wu in #1004
• add agent notebook and documentation by @qingyun-wu in #1052
• Support more azure openai api_type by @thinkall in #1059
• suppress warning message of pandas_on_spark to_spark by @thinkall in #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 #1056
• Fix documentation by @sonichi in #1075
• encode timeout msg in bytes by @sonichi in #1078
• Add pands requirement in benchmark option by @qingyun-wu in #1070
• Fix pyspark tests in workflow by @thinkall in #1071
• Docmentation for agents by @qingyun-wu in #1057
• Links to papers by @sonichi in #1084
• update openai model support by @sonichi in #1082
• string to array by @sonichi in #1086
• Factor out time series-related functionality into a time series Task object by @EgorKraevTransferwise in #989
• An agent implementation of MathChat by @kevin666aa in #1090
• temp solution for joblib 1.3.0 issue by @thinkall in #1100
• support string alg in tune by @skzhang1 in #1093
• update flaml version in MathChat notebook by @kevin666aa in #1095
• doc update by @sonichi in #1089
• Update OptunaSearch by @skzhang1 in #1106
• Support function_call in autogen/agent by @kevin666aa in #1091
• update notebook with new models by @sonichi in #1112
• Enhance Integration with Spark by @levscaut in #1097
• Add Funccall notebook and document by @kevin666aa in #1110
• Update docstring for oai.completion. by @LeoLjl in #1113
• Try to prevent the default AssistantAgent from asking users to modify the code by @sonichi in #1114
• update colab link by @sonichi in #1118
• fix bug in math_user_proxy_agent by @kevin666aa in #1124
• Add log metric by @thinkall in #1125
• Update assistant agent by @sonichi in #1121
• suppress printing data split type by @xiaoboxia in #1126
• change price ratio by @sonichi in #1130
• simplify the initiation of chat by @sonichi in #1131
• Update docs on how to interact with local LLM by @LeoLjl in #1128
• Json config list, agent refactoring and new notebooks by @sonichi in #1133
• unify auto_reply; bug fix in UserProxyAgent; reorg agent hierarchy by @sonichi in #1142
• rename GenericAgent -> ResponsiveAgent by @sonichi in #1146
• Bump semver from 5.7.1 to 5.7.2 in /website by @dependabot in #1119
• autogen.agent -> autogen.agentchat by @sonichi in #1148
• MathChat blog post by @kevin666aa in #1096
• Commenting use_label_encoder - xgboost by @minghao51 in #1122
• raise error when msg is invalid; fix docstr; improve ResponsiveAgent; update doc and packaging; capture ipython output; configurable default reply by @sonichi in #1154
• consecutive auto reply, history, template, group chat, class-specific reply by @sonichi in #1165
• Improve auto reply registration by @sonichi in #1170
• Make auto reply method pluggable by @sonichi in #1177
• support async in agents by @sonichi in #1178
• Updated README.md with installation Link by @royninja in #1180
• Add RetrieveChat by @thinkall in #1158
• silent; code_execution_config; exit; version by @sonichi in #1179

New Contributors

• @badjouras made their first contribution in #1051
• @kevin666aa made their first contribution in #1090
• @LeoLjl made their first contribution in #1113
• @xiaoboxia made their first contribution in #1126
• @minghao51 made their first contribution in #1122

Full Changelog: v1.2.4...v2.0.0

v2.0.0rc5

This version makes auto-reply methods pluggable and supports asynchronous mode in agents. An example of handling data steams is added.
Thanks to @qingyun-wu @ekzhu for laying the foundation and reviewing!

What's Changed

• Make auto reply method pluggable by @sonichi in #1177
• support async in agents by @sonichi in #1178

Full Changelog: v2.0.0rc4...v2.0.0rc5

v2.0.0rc4

This pre-release makes lots of improvements in the agentchat framework. Many new applications are enabled.
Thanks @JieyuZ2 @gagb @thinkall @BeibinLi @ekzhu @LittleLittleCloud @kevin666aa @qingyun-wu @LeoLjl and others for your contributions!

What's Changed

• update colab link by @sonichi in #1118
• fix bug in math_user_proxy_agent by @kevin666aa in #1124
• Add log metric by @thinkall in #1125
• Update assistant agent by @sonichi in #1121
• suppress printing data split type by @xiaoboxia in #1126
• change price ratio by @sonichi in #1130
• simplify the initiation of chat by @sonichi in #1131
• Update docs on how to interact with local LLM by @LeoLjl in #1128
• Json config list, agent refactoring and new notebooks by @sonichi in #1133
• unify auto_reply; bug fix in UserProxyAgent; reorg agent hierarchy by @sonichi in #1142
• rename GenericAgent -> ResponsiveAgent by @sonichi in #1146
• Bump semver from 5.7.1 to 5.7.2 in /website by @dependabot in #1119
• autogen.agent -> autogen.agentchat by @sonichi in #1148
• MathChat blog post by @kevin666aa in #1096
• Commenting use_label_encoder - xgboost by @minghao51 in #1122
• raise error when msg is invalid; fix docstr; improve ResponsiveAgent; update doc and packaging; capture ipython output; configurable default reply by @sonichi in #1154
• consecutive auto reply, history, template, group chat, class-specific reply by @sonichi in #1165
• Improve auto reply registration by @sonichi in #1170

New Contributors

• @xiaoboxia made their first contribution in #1126
• @minghao51 made their first contribution in #1122

Full Changelog: v2.0.0rc3...v2.0.0rc4

v2.0.0rc3

Highlights

Added new OpenAI models' support of functions in agents. Thanks to @kevin666aa, @sonichi and @qingyun-wu
Please find a code example in this notebook: https://github.com/microsoft/FLAML/blob/main/notebook/autogen_agent_function_call.ipynb

What's Changed

• temp solution for joblib 1.3.0 issue by @thinkall in #1100
• support string alg in tune by @skzhang1 in #1093
• update flaml version in MathChat notebook by @kevin666aa in #1095
• doc update by @sonichi in #1089
• Update OptunaSearch by @skzhang1 in #1106
• Support function_call in autogen/agent by @kevin666aa in #1091
• update notebook with new models by @sonichi in #1112
• Enhance Integration with Spark by @levscaut in #1097
• Add Funccall notebook and document by @kevin666aa in #1110
• Update docstring for oai.completion. by @LeoLjl in #1113
• Try to prevent the default AssistantAgent from asking users to modify the code by @sonichi in #1114

New Contributors

• @LeoLjl made their first contribution in #1113

Full Changelog: v2.0.0rc2...v2.0.0rc3

v2.0.0rc2

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

New Contributors

• @kevin666aa made their first contribution in #1090

Full Changelog: 2.0.0rc1...v2.0.0rc2

2.0.0rc1

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 #1048
• Improve messaging in documentation by @sonichi in #1050
• create an automl option to remove unnecessary dependency for autogen and tune by @sonichi in #1007
• docs: 📝 Fix link to installation section in Task-Oriented-AutoML.md by @badjouras in #1051
• doc and test update by @sonichi in #1053
• remove redundant doc and add tutorial by @qingyun-wu in #1004
• add agent notebook and documentation by @qingyun-wu in #1052
• Support more azure openai api_type by @thinkall in #1059
• suppress warning message of pandas_on_spark to_spark by @thinkall in #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 #1056

New Contributors

• @badjouras made their first contribution in #1051

Full Changelog: v1.2.4...2.0.0rc1

v1.2.4

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 #1023
• chat completion check by @sonichi in #1024
• update model of text summarization in test by @liususan091219 in #1030
• Human agent by @qingyun-wu in #1025
• fix of website link by @sonichi in #1042
• Blogpost by @qingyun-wu in #1026
• Update default search space by @Kyoshiin in #1044
• Fix PULL_REQUEST_TEMPLATE and improve test by removing unnecessary environment variable by @thinkall in #1043
• response filter by @sonichi in #1039

New Contributors

• @BeibinLi made their first contribution in #1023
• @Kyoshiin made their first contribution in #1044

Full Changelog: v1.2.3...v1.2.4

v1.2.3

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

New Contributors

• @luckyklyist made their first contribution in #1010
• @garar made their first contribution in #1015

Full Changelog: v1.2.2...v1.2.3

v1.2.2

What's Changed

• update nlp notebook by @liususan091219 in #940
• Blog post for LLM tuning by @sonichi in #986
• fix zerodivision by @liususan091219 in #1000
• extract code from text; solve_problem; request_timeout in config; improve code by @sonichi in #999

Full Changelog: v1.2.1...v1.2.2

v1.2.1

What's Changed

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

New Contributors

• @JaneIllario made their first contribution in #993

Full Changelog: v1.2.0...v1.2.1