README

This repository implements Ragas Metrics on a RAG system with OpenLIT observability tool.

Instructions

In order to run this repo:

Setup Instructions

1. Clone the repository:

$ git clone https://github.com/paulomuraroferreira/Self_correcting_coding_agent.git
$ cd Self_correcting_coding_agent

2. Install Dependencies:

$ pip install -e .

3. Environment Configuration:

Create a .env file and fill the following environment variables:

OPENAI_API_KEY=your_openai_api_key
OTEL_EXPORTER_OTLP_ENDPOINT=http://127.0.0.1:4318

4. Run the Application:

   4.1. Execute the chunking process

   Put the pdfs files in the folder app/data/pdfs

   Run

   $ python app/chunking.py

   4.2. Run fastapi

    $ cd app
    $ uvicorn app:app --reload
   

   4.3 Run OpenLit

    $ git clone git@github.com:openlit/openlit.git
    $ cd openlit
    $ docker compose up -d
   

   4.4 Enter the OpenLit dashboard with the following credential:

    Email: user@openlit.io
    Password: openlituser
   

   4.5. Send the request on the files request.http

   Example:

    ### POST Request to FastAPI
    POST http://127.0.0.1:8000/query
    Accept: application/json
    Content-Type: application/json
   
    {
    "question": "What is QLora?"
    }
   

   4.6 Check the metrics on the OpenLit dashboard: