Gradio-based Multimodal LLM Project

This project implements a multimodal Large Language Model (LLM) using Gradio for the user interface.

Colab

Mostly everything is already setuped in ipynb file, to use with gradio UI. Use it with Google Colab for GPU.

Prerequisites

• Python 3.7+
• pip (Python package manager)

Installation

1. Clone the repository:

git clone https://github.com/bedead/Multimodel-LLM-for-creating-testcases--MILVLG-imp-v1-3b-

cd Multimodel-LLM-for-creating-testcases--MILVLG-imp-v1-3b-

2. Create a virtual environment (optional but recommended):

   python -m venv venv source venv/bin/activate

3. Install the required packages:

   pip install -r requirements.txt

Usage

1. Run the main script: python app.py

2. Open your web browser and go to the URL provided in the terminal (usually http://127.0.0.1:7860).

3. Use the Gradio interface to interact with the multimodal LLM:

• Upload images or enter text as input
• Click the "Submit" button to generate results
• View the model's output in the interface

Troubleshooting

If you encounter any issues, please check the following:

• Ensure all dependencies are correctly installed
• Verify that you have the necessary GPU drivers if using GPU acceleration
• Check the console output for any error messages