* Add tokenizer improvments via Singleton class
linting
* dev build
* Estimation fallback when string exceeds a fixed byte size
* Add notice to tiktoken on backend
* feat: add support for voyage-3-large and voyage-code-3 embedding models
- Add voyage-3-large and voyage-code-3 to VoyageAiOptions dropdown
- Update getMaxEmbeddingLength to support 32k context for new models
- Update .env.example with new model options
* unset env example
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* Reranker WIP
* add cacheing and singleton loading
* Add field to workspaces for vectorSearchMode
Add UI for lancedb to change mode
update all search endpoints to pass in reranker prop if provider can use it
* update hint text
* When reranking, swap score to rerank score
* update optchain
* Add support for Google Generative AI (Gemini) embedder
* Add missing example in docker
Fix UI key elements in options
Add Gemini to data handling section
Patch issues with chunk handling during embedding
* remove dupe in env
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* Add support for gemini authenticated models endpoint
add customModels entry
add un-authed fallback to default listing
separate models by expiermental status
resolves#2866
* add back improved logic for apiVersion decision making
* add writible fields to dev api new workspace endpoint
* lint
* implement validations for workspace model
* update swagger comments
* simplify validations for workspace on frontend and API
* cleanup validations
---------
Co-authored-by: Timothy Carambat <rambat1010@gmail.com>
* wip remove all docs clear vector db on embedder/vector db change
* purge all cached docs and remove docs from workspaces on vectordb/embedder change
* lint
* remove unneeded console log
* remove reset vector stores endpoint and move to server side updateENV with postUpdate check
* reset embed module
* remove unused import
* simplify deletion process
rescoped document deletion to be more general for speed, everything needs to be reset anyway
fixed issue where unembedded docs not in any workspaces, but cached, were not removed
* add back missing readme file
update warning text modals
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* Add vector search API endpoint
* Add missing import
* Modify the data that is returned
* Change similarityThreshold to scoreThreshold
As this is what is actually returned by the search
* Removing logging (oops!)
* chore: regen swagger docs for new endpoint
fix: update function to sanity check values to prevent crashes during search
---------
Co-authored-by: Scott Bowler <scott@dcsdigital.co.uk>
* wip hub connection page fe + backend
* lint
* implement backend for local hub items + placeholder endpoints to fetch hub app data
* fix hebrew translations
* revamp community integration flow
* change sidebar
* Auto import if id in URL param
remove preview in card screen and instead go to import flow
* get user's items + team items from hub + ui improvements to hub settings
* lint
* fix merge conflict
* refresh hook for community items
* add fallback for user items
* Disable bundle items by default on all instances
* remove translations (will complete later)
* loading skeleton
* Make community hub endpoints admin only
show visibility on items
combine import/apply for items to they are event logged for review
* improve middleware and import flow
* community hub ui updates
* Adjust importing process
* community hub to dev
* Add webscraper preload into imported plugins
* add runtime property to plugins
* Fix button status on imported skill change
show alert on skill change
Update markdown type and theme on import of agent skill
* update documentaion paths
* remove unused import
* linting
* review loading state
---------
Co-authored-by: Timothy Carambat <rambat1010@gmail.com>
* exposes `maxConcurrentChunks` parameter for the generic openai embedder through configuration. This allows setting a batch size for endpoints which don't support the default of 500
* Update new field to new UI
make getting to ensure proper type and format
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* togetherai llama 3.2 vision models support
* remove console log
* fix listing to reflect what is on the chart
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* feat: add new model provider: Novita AI
* feat: finished novita AI
* fix: code lint
* remove unneeded logging
* add back log for novita stream not self closing
* Clarify ENV vars for LLM/embedder seperation for future
Patch ENV check for workspace/agent provider
---------
Co-authored-by: Jason <ggbbddjm@gmail.com>
Co-authored-by: shatfield4 <seanhatfield5@gmail.com>
* Fix incorrect json API description.
* small edits and validity checks
* remove console.logs
* unset and recheck changes
---------
Co-authored-by: Adam <phazei@gmail.com>
* thread creation additional params name and slug, with api
* typo fix
* Rebuild openai Swagger docs
Handle validations for fields to prevent invalid field inputs for .new
Enforce sluggification of `slug` to prevent breaking of URL structs
---------
Co-authored-by: abrakadobr <abrakadobr@gmail.com>
Adds support for only the llama3.2 vision models on groq. This comes with many conditionals and nuances to handle as Groqs vision implemention is quite bad right now
* Update OpenAI TTS config to allow a custom BaseURL
* uncheck config file
* break openai generic TTS into its own provider
* add space
* hide TTS on user msg
---------
Co-authored-by: Adam <phazei@gmail.com>
* support generic openai workspace model
* Update UI for free form input for some providers
---------
Co-authored-by: Timothy Carambat <rambat1010@gmail.com>
* set message limit per user
* remove old limit user messages + unused admin page
* fix daily message validation
* refactor message limit input
refactor canSendChat on user to a method on user model
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* Refactor handleDefaultStreamResponseV2 function for better error handling
* run yarn lint
* small error handling changes
* update error handling flow and scope of vars
* add back space
---------
Co-authored-by: Roman <rrojaski@gmail.com>
* refactor agent to add fallback to workspace, then to chat provider/model
* commenting
update logic for bedrock and fireworks fallbacks
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* #2317 Fetch pinned documents once per folder to reduce the number of queries.
* Reorder the lines to keeps const declarations together.
* Add some comments to functions
move pinned document fetch for folder to function
move watched documents per-folder to also function the same
remove unused function in documents model
---------
Co-authored-by: Błażej Owczarczyk <blazeyy@gmail.com>
* Issue #1943: Add support for LLM provider - Fireworks AI
* Update UI selection boxes
Update base AI keys for future embedder support if needed
Add agent capabilites for FireworksAI
* class only return
---------
Co-authored-by: Aaron Van Doren <vandoren96+1@gmail.com>
* patch no text results for milvus chunks
* wrap addDocumentToNamespace in try catch for handling milvus errors
* lint
* revert milvus db changes
* add try catch to handle grpc error from milvus
* ollama: Switch from parallel to sequential chunk embedding
* throw error on empty embeddings
---------
Co-authored-by: John Blomberg <john.jb.blomberg@gmail.com>
* initial commit for chrome extension
* wip browser extension backend
* wip frontend browser extension settings
* fix typo for browserExtension route
* implement verification codes + frontend panel for browser extension keys
* reorganize + state management for all connection states
* implement embed to workspace
* add send page to anythingllm extension option + refactor
* refactor connection string auth + update context menus + organize background.js into models
* popup extension from main app and save if successful
* fix hebrew translation misspelling
* fetch custom logo inside chrome extension
* delete api keys on disconnect of extension
* use correct apiUrl constant in frontend + remove unneeded comments
* remove upload-link endpoint and send inner text html to raw text collector endpoint
* update readme
* fix readme link
* fix readme typo
* update readme
* handle deletion of browser keys with key id and DELETE endpoint
* move event string to constant
* remove tablename and writable fields from BrowserExtensionApiKey backend model
* add border-none to all buttons and inputs for desktop compatibility
* patch prisma injections
* update delete endpoints to delete keys by id
* remove unused prop
* add button to attempt browser extension connection + remove max active keys
* wip multi user mode support
* multi user mode support
* clean up backend + show created by in frotend browser extension page
* show multi user warning message on key creation + hide context menus when no workspaces
* show browser extension options to managers
* small backend changes and refactors
* extension cleanup
* rename submodule
* extension updates & docs
* dev docker build
---------
Co-authored-by: shatfield4 <seanhatfield5@gmail.com>
* Refactor api endpoint chat handler to its own function
remove legacy `chatWithWorkspace` and cleanup `index.js`
* Add `sessionId` in dev API to partition chats logically statelessly
* Enable agent context windows to be accurate per provider:model
* Refactor model mapping to external file
Add token count to document length instead of char-count
refernce promptWindowLimit from AIProvider in central location
* remove unused imports
* force lowercase and no space for new and
edit user modals
* edit account modal validations
* use pattern for form validation + remove validations from edit user
* revert comment deletions
* comment fix
* update validation message
* update regex
allow updating by block name changes to invalid names
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* Revert "Patch ODBC support from missing binary/headers for node-odbc"
This reverts commit 9de6b1cc26.
* Revert "OBDC Support (#1933)"
This reverts commit cd597a361e.
* add possibility to connect to SQL Base by ODBC
---------
Co-authored-by: suchaudn <nicolas.suchaud@legrand.fr>
Co-authored-by: nicho2 <nicho2@laposte.net>
* Add support for GitLab repo collection as well as Github Repo collection
* Refactor for repo collectors to be more compact
---------
Co-authored-by: Emil Rofors <emirof@gmail.com>
* return 400 if user creation failed
The client might not check the error field unless they realize it's a 4xx error instead of 200 success.
* Update index.js
* use one-liner style
---------
Co-authored-by: Timothy Carambat <rambat1010@gmail.com>
* Added Supported Models Free Tier - chat_models.txt
Need to fill in correct Parameter Count.
* Bump perplexity model
closes#1901closes#1900
---------
Co-authored-by: Tim-Hoekstra <135951177+Tim-Hoekstra@users.noreply.github.com>
* bump base image
* Bump mysql
* attestations
* attestations perms
* attestations perms
* fix permissions for attestetions for master push
* test cleanup
* repin base image
* revert base
* patch frontend-clean
* force resolve braces to fixed version
* rebump
* patch llm providers that have manual inputs for model pref
* refactor agent model fallback
update UI to show disabled providers to stop questions about provider limitations
* patch log on startup
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* implement delete message feature
* add new delete chat endpoint that just hides messages from ui and fix perms
* refactor: add delete-event hooks for delete so we dont need to pass so much
refactor: implmentation of delete endpoint PUT
* update animation
linting
remove element on animiation to prevent popin
* linting and cleanup imports
---------
Co-authored-by: timothycarambat <rambat1010@gmail.com>
* implement thread forking feature
* rename thread based on forked message
* refactor bulk message create for thread fork + bump prisma version
* revert prisma version bump
* add todo to bulkCreate function in workspace chats
* cast user input to expected type to prevent prisma injection
* refactor: update order of ops for thread fork
---------
Co-authored-by: Timothy Carambat <rambat1010@gmail.com>
