Week 09 Mini Project - Streamlit App with a Hugging Face Model

This project involves creating a web application using Streamlit and integrating it with a Hugging Face large language model (LLM). The application allows users to generate text based on their input.

Project Requirements

• Create a website using Streamlit
• Connect to an open source LLM (Hugging Face)
• Deploy model via Streamlit or other service (accessible via browser)

Project Implementation

This repository contains a Streamlit web application that uses a Hugging Face model to generate text based on user input. The application allows users to enter a text prompt and generate text based on that prompt using openAI's pre-trained language model, gpt2.

Steps to build a Streamlit app with a Hugging Face model

1. Install necessary libraries: This project requires Streamlit and Transformers. You can install them using pip:

pip install streamlit transformers

2. Create a new Python file for the web app: This will be your main app file. Let's call it app.py

3. Import necessary libraries: You will need to import Streamlit and Transformers in your app.py file:

import streamlit as st
from transformers import pipeline

4. Initialize the Hugging Face model: Use the pipeline function from the transformers library to load the model. For example, if you're using a text generation model:

generator = pipeline('text-generation', model='gpt2')

5. Create Streamlit app: Use Streamlit's functions to create the interface for your app. For example, you can create a text input for the user to enter some text, and a button to generate text. Here's an example:

st.sidebar.title("Input Options")
input_text = st.sidebar.text_input("Enter some text")
generate_button = st.sidebar.button("Generate")

6. Generate text based on user input: When the user clicks the "Generate" button, you can use the Hugging Face model to generate text based on the input. For example:

if generate_button:
    with st.spinner("Generating text..."):
        output_text = generator(input_text)[0]["generated_text"]
    st.subheader("Generated Text:")
    st.write(output_text)

7. Run the Streamlit app: You can run the app using the following command:

streamlit run app.py

8. Deploy the app: You can deploy the app using Streamlit's sharing service or other platforms like Heroku or AWS.