About

This is a simple project with aim of using Rasa(Opensource), LLM API together with browser interface. Basically aim is to build a tool to help users create a story together with AI and a chatbot to help guide along the way.

It uses:

1. RASA opensource as chatbot framework.
2. LLMs in form of Huggingface Inference API.

Setup and Running

Setting up:

1. Clone the repo to local.
2. Create a new python environment.
3. Install rasa.
   1. pip install rasa

Getting Huggingface API credentials:

1. Getting text-completion LLM:
   1. Find any text completion model Huggingface that provides Serverless Inference API.
      1. e.g., mistralai/Mistral-7B-v0.1
   2. In the Deploy option of the model, Inference API(Serverless) will be present.
   3. Click on it and get API_URL.
2. Get Huggingface Access Token:
   1. Settings >> Access Tokens >> New Token (Read Access is enough)

Training the Rasa chatbot:

1. Activate python environment.
2. Train the model: rasa train --domain domain

Run the project:

1. Activate python environment.
2. Set the Huggingface credentials as environment variables:
   1. set LLM_API_KEY=your_api_key
   2. set LLM_API_URL=api_url
3. Run rasa core server: make -f Makefile run-server-debug
4. In another terminal run rasa action server: rasa run actions --debug
5. Open the index.html file, thats inside folder front_end, in some browser.

Usage:

1. Say "hi" in message box.
2. Or directly write some line in the center blank page and hit "Continue Story..."