Chatbot AI Assistant (Context Knownledge Base)

Technology: OpenAi, LlamaIndex, LangChain

Method: Completions (davinci model)

Description:

Chatbot developed with Python and Flask that features conversation with a virtual assistant. This uses a context based conversation and the answers are focused on a local indexation file structured with Llamaindex (GPT-index) lib which converts raw data into a vector dataset (Json) a therefore the assistant will use the dataset to provide answers. It allows to define an initial role and personification.

It make use of LlamaIndex (GPT Index) that provides a central interface to connect LLM’s (large language models) with external data. On this demo is used GPTSimpleVectorIndex to created a vector file from the local documents.

LangChain is a framework built around LLMs. The core of the library is to “chain” together different components to create more advanced use cases around LLMs.

Screenshots

How to run (commands Windows terminal with Python 2.7):

Part One: Prepare Environment

• Define necessary parameters (OpenAi API key, ...) on file 'qa_engine.py'

• Initialize virtual environment and install dependencies, run:

    virtualenv env
    env\Scripts\activate
    pip install flask python-dotenv
    pip install openai
    pip install gpt_index langchain
  

Part Two: Prepare local content

• Add documents to folder "content"

Part Three: Run the app

• Initialize the app:

    flask run
  

• Enter "http://localhost:5000" on browser to interact with app

Changelog

• v0.1
  • initial build