v0.5.0

New features

• Self-consistency
• Multi-agent mode
• Cost tracking
• Human-in-the-loop

What's Changed

• Add support for action dependency through args by @20001LastOrder in #423
• Adapt Agent with Pydantic by @20001LastOrder in #420
• Fix non-termination issue of sherpa by @20001LastOrder in #426
• Multiple improvements over state handling by @20001LastOrder in #431
• Change event logging mechanism by @amirfz in #432
• Chained Actions feature version 1 by @Bubbletea98 in #429
• Feature/link scraper as an action by @Eyobyb in #430
• Refactor error handling in AgentErrorHandler by @Eyobyb in #427
• Add NumberValidationAction for validating numbers in text by @Eyobyb in #428
• Add support to asyncrous actions in Sherpa by @20001LastOrder in #436
• Fix function name by @20001LastOrder in #437
• Merge dev branch to main by @20001LastOrder in #439
• Remove the Slack app folder in the source folder by @20001LastOrder in #440
• Improve loading speed and error handling by @20001LastOrder in #441
• Add support for other callbacks: on_exit, before, on_enter, after by @20001LastOrder in #443
• Add helper funcitons for memory management by @20001LastOrder in #445
• Add support for markdown in documentation by @20001LastOrder in #446
• Make extra dependencies as optional by @20001LastOrder in #444
• Adapt Sherpa to newer version of pydantic and improve asynchronous functions. by @20001LastOrder in #447
• load prompts from json and format prompts by @Eyobyb in #434
• Update dependencies by @20001LastOrder in #448
• Add functionality to put task result status for agent run by @20001LastOrder in #449
• Fix ignoring argument type when parsing action arguments by @20001LastOrder in #451
• passing prompt template through config. by @Eyobyb in #450
• Chat-based policy by @20001LastOrder in #454
• Refactor chat based policy and other policies to use sherpa prompt lo… by @Eyobyb in #456
• Update README.md by @Seti17 in #457
• Refactor prompt template initialization and update prompt loading mec… by @Eyobyb in #458
• Add ability to support stopping exectution by function checker by @20001LastOrder in #460
• Update LLM calls from predict to invoke, fix UnboundLocalError, ConfigDict for Pydantic v2 by @Seti17 in #459
• Fix problem of detecting async functions across Python versions by @20001LastOrder in #462
• Refactor events by @20001LastOrder in #464
• Remove some legacy module and tests by @Seti17 in #463
• improve exception handling by @20001LastOrder in #465
• Enhance prompt handling by adding support for JSON and update tests by @Eyobyb in #466
• Update/doc by @Seti17 in #467
• Add documentation for new event types by @20001LastOrder in #468
• Update docstring for files by @Seti17 in #469
• Refactor shared memory by @20001LastOrder in #472
• Add async tests for actions, agents, and memory modules by @Seti17 in #471
• Add event type helpers function in memory/belief.py by @Seti17 in #470
• Refactor prompt structure in prompts.json to use a list structure wit… by @Eyobyb in #473
• Async agent by @20001LastOrder in #474
• Enhance documentation for prompts package by @Eyobyb in #475
• Refactor: Update llm type annotations to use BaseLanguageModel by @Eyobyb in #476
• Feature/self consistency by @20001LastOrder in #477
• Support weight counting during abstraction by @20001LastOrder in #479
• Make different options available to pass the custom prompt by @Eyobyb in #478
• Example/multi agents by @20001LastOrder in #480
• Feat/self consistency - Concretization by @20001LastOrder in #481
• Add documentation for the multi-agent functionality by @20001LastOrder in #482
• Add MCP agent and actions for GitHub and Linear integration by @Eyobyb in #483
• Fix: handle array content in prompt formatting for text prompts. for … by @amirfz in #484
• fix: rename 'dict' field to 'belief_data' in Belief class to resolve Pydantic warning by @Seti17 in #486
• Fix/architecture diagrams rendering by @Seti17 in #485
• Fix/langchain chroma deprecation warning by @Seti17 in #487
• resolve inheritance diagram rendering in webpage by @Seti17 in #492
• Add dynamic enum handling in response schema by @Eyobyb in #489
• Fix/slack optional integration by @Seti17 in #494
• Feat/list attribute support by @Seti17 in #490
• Add Pydantic validation to BaseVectorDB by @Seti17 in #496
• Feature/enhance cost tracking with automatic calculation and reporting by @Seti17 in #498
• Feat/self consistency by @20001LastOrder in #495
• Update Architecture page by @Seti17 in #501
• Add cost tracking page to doc by @Seti17 in #502

New Contributors

• @Seti17 made their first contribution in #457

Full Changelog: v0.4.0...v0.5.0

v0.4.0 Experimental State Management

What's Changed

• Bump version to v0.3.3 by @github-actions in #412
• feat: Entity validation using actions by @Eyobyb in #407
• Add documents for search_refinement.py and usage method under action by @Bubbletea98 in #416
• Documentation improvement by @20001LastOrder in #419
• Add example to the new state machine architecture by @20001LastOrder in #415
• upgrade pydantic from v1 to v2 by @saminegash in #413

Full Changelog: v0.3.3...v0.4.0

v0.3.3

What's Changed

• Bump version to v0.3.2 by @github-actions in #403
• Feature/upgrade langchain dependency from 0.0332 to 0.2.5 by @saminegash in #394
• State machine Integration by @20001LastOrder in #408
• Add dictionary in the belief and actions associated with it by @20001LastOrder in #410
• Add agent feedback policy by @20001LastOrder in #404

New Contributors

• @github-actions made their first contribution in #403

Full Changelog: v0.3.2...v0.3.3

v0.3.2

What's Changed

• Search Refinement by @Bubbletea98 in #393
• Add workflow to automatically create a pull request for version update by @20001LastOrder in #402

Full Changelog: v0.3.1...v0.3.2

v0.3.1

What's Changed

• Added blueprint arugment by @andytai7 in #389
• Reranking by @20001LastOrder in #385
• Update PromptReconstructor to include llm parameter by @Eyobyb in #383
• addressing #395 by @amirfz in #396
• Add workflow to publish PyPI package by @20001LastOrder in #401

Full Changelog: v0.3.0...v0.3.1

v0.3.0

What's Changed

• Allow multiple vector db usage for shared memory by @Eyobyb in #374
• Add usage description to actions by @Bubbletea98 in #377
• Test dependency by @20001LastOrder in #376
• Add human in the loop for blog writer demo by @KulwantSinghYadav in #379
• update doc for fixing warning by @saminegash in #380
• Refactor the agent and action planner module to be policies by @20001LastOrder in #373
• Blog-Writer User-in-the-Loop Documentation by @artificialfintelligence in #382
• remove scripts folder by @saminegash in #381
• Add Base Retrieval Action and Use Pydantic by @20001LastOrder in #375
• refactor combinedID to be present out of the sherpa package by @Eyobyb in #311
• Input/output folder for blogger by @andytai7 in #384
• change the validation logic with regression validation logic, by @Eyobyb in #357
• Bump version to 0.3.0 by @oshoma in #386
• Improve documentation by @oshoma in #388
• Tidy up formatting by @oshoma in #387
• Fix ReadTheDocs.io build break by @oshoma in #390

New Contributors

• @Bubbletea98 made their first contribution in #377
• @KulwantSinghYadav made their first contribution in #379

Full Changelog: v0.2.2...v0.3.0

v0.2.2

What's Changed

• Fix github action failing by @20001LastOrder in #354
• Entity validation with hydra config by @Eyobyb in #304
• Fix/gha fix by @saminegash in #353
• Improve make and reformat files by @oshoma in #356
• Blog-Writer Feature by @artificialfintelligence in #358
• Action format for citation by @20001LastOrder in #355
• Fix security warnings by @oshoma in #365
• Add missing import by @20001LastOrder in #368
• Update tiktoken version by @20001LastOrder in #367
• Add tutorial for citation validation and fix a bug in it by @20001LastOrder in #371
• Cleanup and bump sherpa_ai package version by @oshoma in #370

Full Changelog: v0.2.1...v0.2.2

v0.2.1

What's Changed

• Bump Sherpa version to 0.2.0 by @20001LastOrder in #319
• change the number extractor to spacy by @Eyobyb in #305
• git hub action for the test by @Eyobyb in #202
• Fix publishing error in 0.2.0 by @20001LastOrder in #331
• Fix bug #332 and update instructions by @artificialfintelligence in #333
• Add how to install spaCy pipeline with Poetry to docs by @artificialfintelligence in #337
• sqlalchemy deprecated issue by @Eyobyb in #323
• removing the open book and updating main readme by @amirfz in #338
• Run offline tests on every push and every pull request by @oshoma in #322
• Add tutorial for PDF reader by @20001LastOrder in #334
• create shared memory with vector db , by @Eyobyb in #251
• fix gsite not working by reformulate the query by @20001LastOrder in #342
• Improve Pull Request template by @oshoma in #345
• Fix/citation validation by @20001LastOrder in #341
• change github action checkout and setup version by @saminegash in #349
• Bump version to 0.2.1 by @20001LastOrder in #350

Full Changelog: v0.2.0...v0.2.1

v0.2.0

New features:

• Complete offline test for Sherpa: #234, #309
• Verification steps: #241, #283, #241
• Agent as configuration: #289, #307

Other changes

• remove pypdf by @Eyobyb in #208
• change type by @Eyobyb in #209
• add back pipdf2 by @Eyobyb in #210
• changing docs configuration file and adding files from llm workshop to the open book by @amirfz in #216
• change pypdf2 to pypdf by @Eyobyb in #223
• added a config for test warings by @saminegash in #222
• onboarding checklist to docs by @amirfz in #219
• Fix typo in github templates by @oshoma in #225
• Flag options by @20001LastOrder in #200
• sqlquery to sqlalchemy by @Eyobyb in #233
• Test Cache for OpenAI API by @20001LastOrder in #234
• Add a markdown template for Pull Requests by @oshoma in #243
• Fix dependency on transformers by @20001LastOrder in #249
• Citation validation by @ChenKua in #241
• add multiple gsite feature by @YujingYang666777 in #239
• Fix test break in question reconstructor test by @oshoma in #254
• Fix test break in test_format_logs_with_thoughts_and_command_succeeds by @oshoma in #255
• Set verbose as the default behavior by @20001LastOrder in #250
• Update docs - Dec 8 workshop content uploaded to opeb book by @amirfz in #259
• add financial sentiment llm reference in docs by @sinbycos in #263
• Update Pull Request template by @oshoma in #258
• Improve readme extraction by @oshoma in #253
• Apply project style conventions to all files by @oshoma in #257
• Move test helper functions to the main library by @20001LastOrder in #268
• Test Utils by @20001LastOrder in #269
• Set QA Agent as the default agent in SlackApp and add configuration for the obsolete agent by @20001LastOrder in #270
• Extract number from text by @Eyobyb in #273
• remove embedding by @Eyobyb in #278
• debugging production by @Eyobyb in #279
• Fix issue with the vector db when using pinecone by @20001LastOrder in #276
• Fix dependency issue when building image by @20001LastOrder in #280
• Feature/GitHub actions by @Eyobyb in #282
• Improve citation validation with parameter tuning by @ChenKua in #283
• number validation by @Eyobyb in #277
• add condition to github extractor to fail gracefully by @saminegash in #284
• Run make to fix code style by @oshoma in #288
• adding autodoc by @20001LastOrder in #291
• Docstring by @ChenKua in #290
• Math agent with PAL by @ChenKua in #292
• 75 percent notifier rebased by @oshoma in #296
• Fix test failing caused by math agent and output parsers by @ChenKua in #297
• Hydra config by @20001LastOrder in #289
• documentation for scrape and database by @Eyobyb in #294
• fix link and file scraper bug by @Eyobyb in #286
• Fix slackbot installation with venv instructions by @artificialfintelligence in #308
• Offline test by @20001LastOrder in #309
• Config demo by @20001LastOrder in #307
• Readthedoc fix by @20001LastOrder in #312
• Remove jar file by @20001LastOrder in #316
• chunksize updated to 3000 from 10000" by @saminegash in #310
• Refactor citations to simplify code and improve docs (closes #303) by @oshoma in #317
• Improve docs by @oshoma in #318

New Contributors

• @sinbycos made their first contribution in #263
• @artificialfintelligence made their first contribution in #308

Full Changelog: v0.1.1...v0.2.0

v0.1.1

What's Changed

• Update Self-Improving LLMs section by @gordania in #1
• changes to SP talk by @piesauce in #2
• Update README.md by @stillonearth in #3
• Add TOC Generator by @ianyu93 in #4
• Added link for the slides by @SinaShahandeh in #5
• Edit summary and add link to the notebooks by @20001LastOrder in #6
• Adding files for slackbot by @babasahebpinjar in #10
• Adding link to create slack app by @babasahebpinjar in #11
• Integrate Slack App with Pinecone by @20001LastOrder in #12
• React Slack Bot by @20001LastOrder in #16
• Citation fix and local sandbox by @20001LastOrder in #17
• Update FT slides link by @ezzini in #30
• Update bolt_app.py for verbose message output by @johnnykfeng in #41
• modified task_agent.py to handle logging messages by @johnnykfeng in #40
• Update and rename Reflexion: A Reflection-Based Approach.rst to Refle… by @johnnykfeng in #43
• Update and rename TempLM: LM Distilling Language Models into Template… by @johnnykfeng in #44
• Update and rename Auto GPT: A Breakthrough in Language Models and Mac… by @johnnykfeng in #45
• Slackbot tutorial by @20001LastOrder in #47
• Separate verbose message in Slack. Added -verbosex flag. by @johnnykfeng in #51
• remove old file to avoid confusion by @20001LastOrder in #55
• Create transcript_summarizer_aws_lambda.py by @johnnykfeng in #60
• Update prompt_generator.py by @bharatji30 in #62
• add .env to gitignore by @saminegash in #65
• update book index by @amirfz in #71
• Add linter and formater to the project by @20001LastOrder in #72
• key error resolved by @saminegash in #74
• convert makrdown link to slack link by @20001LastOrder in #66
• link , scraper for question by @Eyobyb in #68
• Updating docs by @amirfz in #85
• adding requirement.txt by @amirfz in #86
• Update requirements.txt by @amirfz in #87
• Fix app startup defects by @oshoma in #84
• Converted the file normal links to permalinks by @aminfadaei116 in #91
• Solve #38 none command and #23 error handeling by @YujingYang666777 in #63
• reformulate query if the query is duplicate by @YujingYang666777 in #69
• Add draft GitHub workflow guidelines by @oshoma in #94
• Rename repository from practical-llms to sherpa by @oshoma in #97
• fix url content extraction by @saminegash in #95
• App initialization by @oshoma in #89
• Fix the ordering of initializing the flask app by @20001LastOrder in #100
• Update issue templates by @amirfz in #102
• Removed redundant call by @hakeemo in #99
• reordering the function definition above the call by @saminegash in #108
• Update LICENSE by @amirfz in #103
• Update issue templates by @amirfz in #107
• Enable debugging mode for local development (v2) by @oshoma in #106
• Serper api added to the config by @saminegash in #109
• Refactoring of task agent and configure logger by @20001LastOrder in #88
• add pinecone_setup.md in docs/Tutorials by @johnnykfeng in #112
• Fix typo in search tool description by @oshoma in #118
• Format slackbot files v2 by @oshoma in #117
• Improve logging v2 by @oshoma in #120
• updating the doc structure by @amirfz in #96
• add reflection class by @YujingYang666777 in #122
• updating some docs referencing the old docs URL by @amirfz in #128
• changing credential for docker plus adding retrieval from aws secret … by @saminegash in #129
• change the temprature to 0 by @Eyobyb in #131
• add log for the username and pwd" by @saminegash in #132
• credential check by @saminegash in #134
• fix retrieval by @saminegash in #135
• fix username and password problem by @saminegash in #136
• update Pinecone script by @ChenKua in #139
• create verbose logger template by @20001LastOrder in #140
• Resource Citation by @20001LastOrder in #83
• Update task_agent.py for typo and missing words by @bharatji30 in #142
• protect key access by @ChenKua in #144
• Error handling for agent task execution by @20001LastOrder in #141
• include verbose message at every step by @johnnykfeng in #147
• Remove aggregated verbose message at the end of the task by @20001LastOrder in #148
• usage Tracker and whitelist by @Eyobyb in #143
• Description about how to access configuration variable values by @Aminolroaya in #145
• small changes to verbose message format by @johnnykfeng in #150
• Local Vectordb with Production Data by @20001LastOrder in #146
• Refactoring Part 1 by @20001LastOrder in #152
• Add instructions to run the refactored project by @20001LastOrder in #157
• fix typo and add whitelist endpoint by @Eyobyb in #161
• Skip daily limit check in debug mode by @20001LastOrder in #162
• clean up refactoring and fix a typo in instruction by @20001LastOrder in #163
• Requirement to run docker-compose by @Aminolroaya in #159
• build error fix by @saminegash in #166
• build error fix by @saminegash in #167
• build error fix by @saminegash in #168
• build error fix by @saminegash in #169
• Revert changes in dockerfile to fix build by @20001LastOrder in #171
• Add pytest-cov for test coverage and add integration test by @20001LastOrder in #164
• add description to context tools by @YujingYang666777 in #165
• add a count down function by @Eyobyb in #173
• fix current time issue by @Eyobyb in #175
• Set default react iteration to 1 by @20001LastOrder in #178
• Ask user for more context to finish a task by @hakeemo in #160
• add youtube link in How_To docs by @johnnykfeng in https://github.com/Aggregate-Intellec...