1.23.0-beta

Remove crew, replace with custom tool-calling (#243)

* wip on moving away from crew

* wip

lastFile:llm-service/app/services/query/agents/crewai_querier.py

* drop databases

lastFile:llm-service/app/services/query/agents/crewai_querier.py

* WIP basic tool calling and minor refactoring of retriever tool

lastFile:llm-service/app/services/query/tools/retriever.py

* not working usage of openai agent

* basic streaming example

* wip on openai tool using

* provide summaries to the retrieval tool

* added source nodes to OpenAI agent stream

* minor change to prompt

* wip on supporting bedrock

* WIP streaming for non open ai models

* add conditional for openai

* support for non openai agents and refactor

* minor fix to return types

* remove crew

* minor change to check for empty response

* fix mypy, ruff issues

* Revert "bump mui packages that seemed to address issues with x-charts"

This reverts commit 864b770f611ee0e4d85e44b2b2dedbe34e83e3d7.

* Revert "update hide legend"

This reverts commit b4db0dde75ab3235345fa2452cff4e8d187a00db.

* remove the last crew bits

* added verbosity for non openai agents

* added missing print dashes

* better error handling for bedrock model availability

* fix mypy

* fix writing source nodes

---------

Co-authored-by: Elijah Williams <ewilliams@cloudera.com>
Co-authored-by: Michael Liu <mliu@cloudera.com>
Co-authored-by: Baasit Sharief <baasitsharief@gmail.com>

1.22.2

add branch (#242)

hotifx

add branch (#242)

1.22.1

kill the startup script if python fails to start (#237)

* kill the startup script if python fails to start

* Update scripts/startup_app.sh

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

1.22.0

a few tweaks for better logging & old CML/crew support (#231)

* a few tweaks for better logging & old CML/crew support

* add pysqllite3 as a dep for Linux

* add pysqllite3-binary to the deps, and put the import into a try/except

* ruff

* print out the system platform for debugging purposes

* Update llm-service/app/main.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* address deprecation issues

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

1.22.0-dev

add some simple alerts when a model does not exist (#229)

* add some simple alerts when a model does not exist

* add nav buttons

* switch to using suspense query for amp config

* better error handling for streaming

* bumping crewai configs to be more lenient

* bump max iter time to 120 seconds

* fixed merge conflicts

* bump back to 120

* use the proper model name for the crew planner agent

* changed to runtime error

* unquote crew ai errors

* added more info in tips

* remove unused import

* Update llm-service/app/services/query/agents/crewai_querier.py

Co-authored-by: mliu-cloudera <mliu@cloudera.com>

* remove unquote

---------

Co-authored-by: Baasit Sharief <baasitsharief@gmail.com>
Co-authored-by: jwatson <jkwatson@gmail.com>
Co-authored-by: mliu-cloudera <mliu@cloudera.com>

1.22.0-beta

OpenSearch support (#223)

* add http auth for the low level client as well

* wip on chunk retrieval

* implement raw query to return the node contents for OpenSearch

* don't refetch chat history on window focus to allow for continued streaming

* add in oss username and password

* drop databases

lastFile:llm-service/app/services/amp_metadata/__init__.py

* create opensearch config for ui and backend

* wip

lastFile:ui/src/pages/Settings/VectorDBFields.tsx

* drop databases

lastFile:ui/src/pages/Settings/VectorDBFields.tsx

* WIP vector db provider

lastFile:ui/src/pages/Settings/VectorDBFields.tsx

* cleanup opensearch config setting

* add namespace for opensearch

* Handle 404s properly with opensearch deletions & other possibly failed deletion

* fix a small mypy issue

* add beta tag

* Update ui/src/pages/RagChatTab/SessionsSidebar/CreateSession/CreateSessionForm.tsx

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* handle the case where we can't find the chunk

* use typeadapters

* log query

---------

Co-authored-by: Elijah Williams <ewilliams@cloudera.com>
Co-authored-by: Michael Liu <mliu@cloudera.com>
Co-authored-by: Baasit Sharief <baasitsharief@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

1.21.0

Merge branch 'release/1' into main

1.21.0-beta

Tool Calling (#219)

* feat: added tool calling with retrieval and search results (#214)

* Retrieval and Searching happens in tandem
* Source nodes cited but not available in chat response are also provided to chat response with a score of 0.0 i.e. source node information used from chat history
* Reduced llm calls by removing the number of agents and using only tasks
* Further refactoring might be needed

* support openai for tool calling

* Propagate error from chat streaming

* fix mypy issues

* conditionally do evals based on whether we have source nodes

* don't show direct llm call alert if tool calling is enabled

* add opik for llama-index, fix chat history query!

* catch an error when trying to extract source nodes from responses

* don't scroll to top of response when done streaming

* added azure api version to settings and prompt changes

* wip on using mcp with crewai

* switch to playing with the fetch mcp server

* remove the unneeded env to the uvx command

* refactor to allow multiple MCP servers/toolsets, and add the mcp-server-fetch as an option

* add in text2sql2text tool

* move to using mcp.json

* add selectedTools as an attribute of the session queryConfiguration

* hook up selected tools to the UI

* use the sessions's tools settings

* rename amp settings to studio settings

* remove path to caii domain

* fix table name

* remove tools from query configuration and remove calls to search (leave in search code for now)

* move mcp.json file to tools in root

* fix mypy

* Update release version to dev-testing

* use the right model name for crew

* added verbosity back

* Add some text to the tool calling options abot the power of the inference model

* readme for the mcp.json

* remove text2sql tool

* minor changes to description and tools manager formatting

* add support for env

* add in the root environment to the tools

* add tools/ to gitignore

* load node env before starting up python

* fix a mypy issue

* fix mypy issue with missing model

* remove console.error

* bug fix to passing task contexts, and minor prompt changes

* remove serper tool

* Rename model provider tests

* fix bug with overriding tools when updating session

* added date to researcher task description

* Rename methods; use list instead of List

* Add methods to provider interface

* add more logs for tool calling

* Implement methods

* Clean up code

* wip on moving chat history

* get ruff passing

* WIP mcp tool calling

* ruff check

* remove mcp server

* add beta tags

* remove comment on chat history tool

---------

Co-authored-by: Baasit Sharief <baasitsharief@gmail.com>
Co-authored-by: jwatson <jkwatson@gmail.com>
Co-authored-by: Michael Liu <mliu@cloudera.com>
Co-authored-by: actions-user <actions@github.com>

1.20.0

remove the required asterisk from the base url & fix cross-config pro…