Skip to content

Releases: mindsdb/mindsdb

v24.5.4.0

22 May 11:21
5f4cb98
Compare
Choose a tag to compare

What's Changed

Read more

v24.4.3.0

15 Apr 13:43
d097d20
Compare
Choose a tag to compare

Major updates of this release include the integration of MindsDB with MindsDB Inference Endpoints, which is a service provided by MindsDB that lets you access various LLMs with a single API key. Besides, MindsDB has been working on updates to AI Agents, including knowledge bases and RAG features.

New handlers:

  • Mindsdb inference endpoints #9044

Features, Improvements and Fixes:

  • bump litellm version #9064
  • Managing knowledge base columns #9005
  • Integrate rag with langchain agents #9027
  • Initial RAG 2.0 integration #9068
  • Option to disable storage backup #9073
  • Kb updates 04.2024 #9069

Documentation Updates:

v24.4.2.1

11 Apr 16:49
8c4f890
Compare
Choose a tag to compare

Features, Improvements and Fixes:

  • Add RAG model and utilities for Mindsdb integration #8996
  • Update requirements.txt #9048
  • Update requirements.txt #9050
  • update sampledatabases to new k8s server #8986
  • The "database" field in the form is changed to not required. #9052
  • Fixed Column Structure of chat_messages Table #9015
  • Fix BYOM engine create #9054
  • Clear LAST value in editor #9049

Documentation Updates:

  • updated agents docs #9058
  • updated broken links in mongo docs #9059
  • Feature/update dd docs #9062

v24.4.2.0

10 Apr 03:37
d767b23
Compare
Choose a tag to compare

What's Changed

New Contributors

Full Changelog: v24.3.5.0...v24.4.2.0

v24.3.5.0

26 Mar 10:44
b946e2a
Compare
Choose a tag to compare

Features, Improvements and Fixes:

  • create prometheus dir if it doesnt exist #8991
  • add default handlers to manifest and fix pip install test #8989
  • agent behaves as model #8984
  • Fix image tags #8979
  • Add Local Ollama Model Support to Langchain Handler #8978
  • Sets the 'success' Attribute of /<handler_name>/install Endpoint to False Upon Installation Failure #8977
  • Add Support for Querying Agents via SQL #8856
  • New /tabs routes #8898
  • added ollama_serve_url argument to overwrite hardcoded default URL and validated linux compatability #8914
  • Updated the Broken Dependencies for the BigQuery Handler #8910
  • Github handler: all tables converted to APIResource #8940
  • Del stream interface #8967

Documentation Updates:

  • updated mindsdb editor image in docs #8982
  • mindsdb sql updates for docs #8939
  • updated docker docs #8938

v24.3.4.2

22 Mar 12:14
9fa25bd
Compare
Choose a tag to compare
  • fixed docker image tag

v24.3.4.0

20 Mar 16:17
39ca223
Compare
Choose a tag to compare

Features, Improvements and Fixes:

  • [ENH] Code and contextual QA formatters for LLM finetuning #8759
  • Fixes for Issue #8565 #8567
  • Refactor email handler #8794
  • clean up docker/docker-compose and comment #8798
  • Add 12labs handler to cloud #8804
  • Improve email ingestion and handling in MindsDB #8803
  • Added insert method for email campaigns table #8698
  • Skills updates #8765
  • Refactor docker bake file #8763
  • Improve error message if model in 'error' status #8827
  • Fix create ml engine #8830
  • Fix describe of broken openai model #8829
  • Fix: JOIN model with empty dataframe #8843
  • PR for small fixes on hachaton #8833
  • Fix describe for models with 'broken' ml_engines #8839
  • Validate api key during OpenAI or Anyscale-endpoints ml_engine creation #8847
  • Validate the timegpt token #8841
  • Add multi environment deploy step #8836
  • add deploy to staging when mergeing to staging branch #8852
  • Fix show command #8849
  • Adding icon.svg file for writer handler #8859
  • Removed Non-integrations from the /handlers Endpoint #8862
  • Fixed Several Unusable Handlers #8864
  • Added Support for Summarization to the Twelve Labs Integration #8795
  • Make Langchain Handler More Reliable #8844
  • Add /completions POST Endpoint for Agent Completions #8850
  • [ENH] Updates to langchain: version bump, observability #8785
  • Getting anyscale key instead openai #8800
  • Skills decoupling #8792
  • Updated icons for TPOT and StatsForecast #8872
  • Slack list messages updated #8796
  • Don't pickle exception, wrap it #8866
  • Make langchain_community dependency optional #8875
  • Mark files handler as permanent #8879
  • fix github_tables filtering #8881
  • Make web handler optional #8880
  • Twelvelabs Integration fixes #8883
  • Fixed the Import of MatrixOneHandler in D0lt Handler #8886
  • Fix last: one condition in view #8867
  • Add auth flag in the config example #8888
  • Anomaly detection process data bug #8897
  • Fix warning about flask JSONEncoder deprecation #8869
  • Init load tests #8730
  • anthropic to use messages api (instead of legacy completions api) #8885
  • Removed Icons as Determining Factor for Handler Import Status #8909
  • Updated the Broken Dependencies for the BigQuery Handler #8910
  • Add Support for Prometheus Metrics & Add New Handler Metrics #8899
  • fix for email handler to work in jobs with last #8913
  • Fix failing coveralls action #8917
  • Fix select connection_id() #8920
  • Added Improved Logos for Integrations #8854
  • Add jobs_history to log db #8793
  • Bump follow-redirects from 1.15.5 to 1.15.6 in /docs #8932
  • Fix supported python versions for pip installs #8905
  • Json operators support #8895
  • fix setup limits #8937
  • Add Prometheus Metrics for All API Endpoints #8904
  • Include prometheus_metrics Dir in Dockerfile #8944

Documentation Updates:

  • Typo "mbd" vs "mdb" #8736
  • Update README.md - tweak messaging #8797
  • [Docs] update mongo doc: replace insert->insertOne, predictors->`models #8802
  • fixed api_key param names #8809
  • added info to finetune commands #8808
  • docs for evaluate by #8807
  • update to rest api docs #8810
  • docs rearranged and modified intro pages #8811
  • Cleanup docs #8825
  • added a note that create table needs database with write access #8832
  • Update Anomaly detection docs #8834
  • added eeg forecast tutorial #8835
  • tutorial for learning hub #8837
  • fixed query in tutorial #8840
  • updated tutorials for learning hub #8845
  • added data file to openai finetuning tutorial #8853
  • update file link #8855
  • fixed slack chatbot tutorial #8868
  • Less redundant doc example #8873
  • fixed github docs #8877
  • fixed anthropic model names in docs #8884
  • Update README.md #8887
  • Update README.md #8889
  • Update monkeylearn.mdx #8891
  • Update README.md #8901
  • updated middleware to platform #8894
  • added links to use cases #8912
  • update to docker docs #8916
  • updated slack docs #8921
  • Update MindsDB Installation for Development docs #8908
  • updating ai engines doc pages #8799

v24.2.3.0

19 Feb 15:20
bd1b230
Compare
Choose a tag to compare

Features, Improvements and Fixes:

  • Fix args in test_predict_text_sentiment_openai_tutorial #8789
  • Better error handling if exception can't be pickled #8729
  • added rag handler param #8781
  • Slack fixex #8782
  • Fix getting api key #8786
  • Add internal log database #8778
  • Fix All Tests Under tests/unit #8662
  • Stored MS Teams Redirect URI in Handler Storage #8766
  • Corrected Use of Logger in LibSQL Handler #8769
  • modify get_api_key util function and affected handlers #8245
  • Datanode returns list of lists instead of dict #8745
  • Fix getting api-key for openai finetune #8762
  • Fix Twelve Labs indexing local file #8748
  • [ENH] Improved data preparation for LLM finetuning #8692
  • Mongodb handler tests #8470
  • Save anomaly_detection files to storage #8724
  • Minor Fix to the Put Config API Endpoint #8732
  • Update APITable docs #8738
  • Remove permanent for some ml handlers #8737
  • Google_analytics is not permanent #8734
  • [refactor] Migrate legacy OpenAI API #8541
  • [fix] Minor finetune improvements #8727
  • Show content in files by default #8731
  • GITHUB handler update #8721
  • Fix anyscale-endpoints finetune #8713
  • Pass info about model target to planner #8722
  • Postgres handler unit tests #8214
  • Mssql handler unit tests #8317
  • Mysql handler tests #8587
  • Move lanchain skills to langchain handler. #8711
  • Fixed: Whatsapp logger issue #8717
  • Fix RAG post openai changes #8674
  • Run all ML handlers methods via queue #8645
  • Return tests with last to #8662 #8694
  • [fix] Key error when exiting context handler #8710
  • Move file test to acceptance tests #8697
  • Fix mongodb parsing #8675
  • added text-generation support to hf handler #8513
  • PostgreSQL Integration improvments #8680
  • [fix] Anyscale key getter #8693

Documentation Updates:

  • Update Other Clouds Deployment Page #8788
  • rag param update in docs #8780
  • added anyscale docs #8779
  • updated last keyword docs #8776
  • updated intro docs #8775
  • updated slack chatbot tutorial #8772
  • Update twitter-chatbot.mdx to use LAST #8750
  • updated mindsdb intro #8770
  • Update README.md with the new messaging #8768
  • updates to integrations doc pages #8758
  • updated neuralforecast parameters in docs #7336
  • sql api updates #8756
  • updated deprecated openai models #8755
  • added d0lt.mdx #5383
  • youtube docs for connecting #8744
  • youtube docs #8741
  • updates to postgres docs #8740
  • updated ai tables doc page #8706
  • installation docs updates #8705
  • PostgreSQL Updates docs #8707
  • youtube documentation update #8682
  • Fix: Typo in documentation #8703
  • sql syntax document page #8676
  • fix_query #8683

v24.1.4.0

25 Jan 13:35
586caf7
Compare
Choose a tag to compare

New Handlers:

  • Whatsapp Integration #8393
  • Initial Google Analytics Handler #8586
  • Leonardo AI Integration #8553
  • Microsoft Teams Chat Handler #8479
  • Add LiteLLM handler #8507

Features, Improvements and Fixes

  • Added dependencies to k8s, LLM fixes #8652
  • Update of internal DB #8552
  • updated sendinblue_tables.py #8679
  • Allow to use in in show tables {from | in} db #8671
  • add credentials_json #8635
  • Fix BYOM venv paths on windows #8668
  • Mark anyscale endpoints integration as permanent #8654
  • Redirect URI Fix for MS Teams #8659
  • fix drop table for vectordatabase_handler.py and specifically chroma #8612
  • Add QueryStep to planner #8524
  • Ollama - Drop model registry check & improve description #8631
  • Remove mssql as default handler #8647
  • YouTube Integration Fixes #8614
  • fix openai handler api_base argument #8630
  • Minor Fixes to LangChain and Rag #8628
  • More deploy triggers and simplify docker #8609
  • Fix vertex dependency #8620
  • Fix deploy trigger input name #8616
  • Fix Replicate schema bug #8601
  • Executor moved from mysql server #8544
  • Fix neuralforecast and statsforecast in cloud image #8605
  • Fix https CH bug #8597
  • fix issue with pdf upload in cloud #8589
  • Bump lightwood to 23.12.4.0 #8602
  • Fix checking frontend version #8598
  • Update LlamaIndex handler with GithubLoader and YoutubeTranscript loader #8562
  • Update deps install log message #8571
  • Feat add new handler to update Shopify orders #8585
  • Update mysql_handler.py #8575
  • Twelve Labs Fixes & Improvements #8578

Documentation Updates

  • updated docs re agents/chatbots/kb #8660
  • Binance #7088
  • Readme updates #8661
  • updated timegpt docs with more info #8657
  • removed chatbot ui docs #8655
  • updated agent/chatbot tutorial #8656
  • restructured docs sections into tabs #8633
  • Add Python code docs for Contributors #8651
  • Update jobs.mdx #8646
  • Update timegpt.mdx #8634
  • Update youtube.mdx #8621
  • Update openai.mdx #8625
  • Update README.md #8622
  • README.md update #8627
  • Fix vertex install instructions #8626
  • updated syntax of describe, retrain, finetune #8618
  • linked finetuning blog post #8617
  • added info about which slack connection method to use with the chatbot syntax #8615
  • Update lightwood documentation URL #8604
  • Update default project #8607
  • Replicate docs update #8599
  • updated timegpt docs #8596
  • new finetuning example for anyscale_endpoints engine #8581
  • Minor docs update #8545
  • fixed tab content in docs #8580
  • Docs updates [fixes to broken links and some docs] #8583

v23.12.4.2

05 Jan 10:56
5f74964
Compare
Choose a tag to compare

Features, Improvements and Fixes

  • migrate openai version #8541
  • fix anyscale issues with finetuning #8579
  • Implemented Components table for instatus handler #8339
  • Update LlamaIndex handler with GithubLoader and YoutubeTranscript loader #8562