Closed jared-yu-hcltech closed 2 months ago
Currently working on a high-level, but having some minor issues. Had to edit the schema to include the LLM. This impacted both old code and new code. Currently it only seems to work smoothly for gpt-4o, the Gemini version seems to not be working. Also, need to sync the Change Model buttons, we probably want to disable them in the case of it being a Custom Chatbot.
The feature has now been largely implemented via c8f7350.
Going to close for now as the feature seems largely functional and file a bug for the remaining issue.
Edit: Apparently the bug has been fixed with 77c3e29. Will not be filing a bug issue then.
In the new chat page, make a clickable button for users to create their own chatbots that they can then communicate with, where they can select their own model and create prompts for it.
Goal: Users can select from the side menu to create their own chatbot. From there, they can submit a prompt which will be used as the system prompt for the LLM. The user will also be allowed to set their own LLM (e.g., gpt-4o, gemini-flash-1.5, etc.).
This will initialize a chat history which then is visible in the normal DashboardPage, and will continue from there. Then, if we leave the chat and come back, we need to make sure that the custom prompts are still properly in effect. We also probably will want to lock it so that the user can't change the model in the middle of the chat for these custom chatbots.