🎥 Text-to-Video Generator Streamlit App

This is a simple yet powerful web application built with Streamlit that allows you to generate short videos from text prompts using the state-of-the-art Damo-VILAB Text-to-Video Diffusion Model.

✨ Features

• Text-to-Video Generation: Convert descriptive text prompts into dynamic video clips.

• Intuitive UI: Easy-to-use interface powered by Streamlit.

• Video Playback: Directly view generated videos within the application.

• Download Option: Download your generated videos for offline use.

• GPU Accelerated: Leverages CUDA for faster video generation (requires compatible NVIDIA GPU).

🚀 Prerequisites

Before you begin, ensure you have the following installed:

• Python 3.8+: Recommended version.

• pip: Python package installer.

• NVIDIA GPU (CUDA Compatible): Highly recommended for performance. The model runs on torch.float16 and is moved to "cuda", so a GPU is essential for practical use. Without a GPU, generation will be extremely slow or may not work due to memory constraints.

🛠️ Installation

Follow these steps to set up and run the application locally:

1. Clone the Repository (or save the code):

   If your code is part of a larger project, navigate to your project directory. If it's a standalone file, save it as app.py.

2. Create a Virtual Environment (Recommended):

   It's good practice to use a virtual environment to manage dependencies.

   python -m venv venv
   venv\Scripts\activate  # On Windows

3. Install Dependencies:

   Install the required libraries using pip. The torch installation specifically points to a CUDA-enabled version. Adjust cu118 to match your CUDA version if different (e.g., cu121).

   pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 --index-url https://download.pytorch.org/whl/cu118
   pip install diffusers accelerate streamlit

   • Note on torch: If you do not have a CUDA-compatible GPU or encounter installation issues, you might need to install a CPU-only version of PyTorch, but expect significantly slower performance. Refer to the PyTorch website for specific installation commands based on your system.

💡 Usage

1. Run the Streamlit Application:

   Navigate to the directory containing your main.py file in your terminal and execute:

   streamlit run main.py

2. Access the App:

   Your web browser will automatically open to the Streamlit application (usually at http://localhost:8501).

3. Generate Videos:

   • Enter your desired text prompt in the input box.

   • Click the "Generate Video" button.

   • The app will display a "Generating video..." message. This process can take several minutes depending on your GPU, internet speed (for model download), and the prompt complexity.

   • Once generated, the video will appear on the page, and a "Download Video" button will be available.

⚠️ Important Considerations

• Performance: Video generation is computationally intensive. A powerful NVIDIA GPU with sufficient VRAM (e.g., 8GB+) is strongly recommended for reasonable generation times.

• First Run: The first time you run the application, the model will be downloaded, which can take a while depending on your internet connection. Subsequent runs will use the cached model.

• Video Length: The current setup generates a fixed number of frames per iteration (num_iterations = 4). You can experiment with this value in the code, but keep in mind that more frames will increase generation time and VRAM usage.