* style(FunctionsAgent.js): remove unnecessary comments and update PREFIX variable
refactor(initializeFunctionsAgent.js): update to use initializeAgentExecutorWithOptions
deps(package.json): update langchain to v0.0.95
refactor(askGPTPlugins.js): pass endpointOption to onStart function
* fix(ChatAgent.js): handle undefined delta content in progressMessage.choices array
* style(client): adjust height and add overflow to EditPresetDialog, AgentSettings, and Settings components
style(client): adjust size and add overflow to NewConversationMenu and PresetItems components
style(ui): add overflow to DialogContent component
* style(Settings.jsx): change height of settings component to md:h-[350px] h-[490px]
* chore(docker): add .env and **/.env to .dockerignore
refactor(docker): remove unnecessary .env file copy and removal in Dockerfile
* style(AgentSettings): adjust Switch placement fix(EditPresetDialog): correctly show functions setting in preset
feat(ChatAgent.js): add a check for images when completion is skipped to add to response
feat(askGPTPlugins.js): add skipCompletion option to agentOptions
feat(client): add Switch component to ui components and use for new Agent Settings
chore(package.json): ignore client directory in nodemonConfig
The `toolInput` object was not being properly logged in the `ChatAgent.js` and `handlers.js` files. The `JSON.stringify()` method was added to properly log the object.
- Improve logging format in ChatAgent.js by adding more details to the log
- Add support for functionsAgent in ChatAgent.js to format the log differently
- Improve formatAction function in handlers.js to handle empty thoughts and add support for functionsAgent
Add support for selecting an agent in the GPT plugins endpoint. The agent can be selected from a dropdown menu in the AgentSettings component. The default agent is set to 'classic' in the cleanupPreset, getDefaultConversation, and handleSubmit functions.
* feat: add api for config
* feat: add data service to client
* feat: update client pages with values from config endpoint
* test: update tests
* Update configurations and documentation to remove VITE_SHOW_GOOGLE_LOGIN_OPTION and change VITE_APP_TITLE to APP_TITLE
* include APP_TITLE with startup config
* Add test for new route
* update backend-review pipeline
* comment out test until we can figure out testing routes in CI
* update: .env.example
---------
Co-authored-by: fuegovic <32828263+fuegovic@users.noreply.github.com>
* feat(OpenAI, PaLM): add new models
refactor(chatgpt-client.js): use object to map max tokens for each model
refactor(askChatGPTBrowser.js, askGPTPlugins.js, askOpenAI.js): comment out unused function calls and error handling
feat(askGoogle.js): add support for codechat-bison model
refactor(endpoints.js): add gpt-4-0613 and gpt-3.5-turbo-16k to available models for OpenAI and GPT plugins
refactor(EditPresetDialog.jsx): hide examples for codechat-bison model in google endpoint
style(EndpointOptionsPopover.jsx): add cn utility function import and use it to set additionalButton className
refactor(Google/Settings.jsx): conditionally render custom name and prompt prefix fields based on model type
The code has been refactored to conditionally render the custom name and prompt prefix fields based on the type of model selected. If the model starts with 'codechat-', the fields will not be rendered.
refactor(Settings.jsx): remove duplicated code and wrap a section in a conditional statement based on a variable
style(Input): add z-index to Input component to fix overlapping issue
feat(GoogleOptions): disable Examples button when model starts with 'codechat-' prefix
* feat(.env.example, endpoints.js): add PLUGIN_MODELS environment variable and use it to get plugin models in endpoints.js
* Create HetznerUbuntuSetup.md
Step-by-step guide for someone who is starting from scratch on this project with a bare server.
* Updated Readme & Heroku
I submitted the original Heroku.md (to the discord) and they are way out of date. Just corrected them, moved the Hetzner file to the cloud deploy, and updated the readme to point to the file.
* Update HetznerUbuntuSetup.md
* Update README.md
- Remove api/.env and .env from .dockerignore file
- Add COPY .env .env to Dockerfile to copy .env file to docker build
- Add RUN rm .env to Dockerfile to remove .env file after build
- Remove build args from docker-compose.yml file
