v24.3.4.0

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