A personalized quiz system using retrieval augmented generation.
First, clone this repository to your local machine:
git clone https://github.com/simoncwang/RAGQuiz.git
Then, create a Python environment
Conda:
conda create -n rag-quiz python=3.12
activate the environment with
Conda:
conda activate rag-quiz
Next, install the required packages with:
pip install -r requirements.txt
Finally, launch the app by running:
python3 app.py
which should launch a Gradio demo, follow the url given to the web app.
Below is a screenshot example of the UI, once running it will contain some instructions to use the app! (detailed instructions will be updated in the future)
NOTE: You can submit multiple files to be stored in the system, and the generate quiz function will randomly select a file from your list to create a question! Files can be uploaded directly from your computer, OR through URL (content is scraped using Selenium).
By default, the app.py script launches a demo running on a local server. If you would like to create a share-able public link, on the last line of app.py add the "share=True" argument to the Gradio demo launch. This will create a link that lasts for 72 hours as long as you leave the script running, and can be shared with others.
demo.launch(server_name='0.0.0.0', share=True)