Skip to content

Latest commit

 

History

History
88 lines (62 loc) · 3.29 KB

README.md

File metadata and controls

88 lines (62 loc) · 3.29 KB
Raw

PixArt Image Generation Project

Overview

This project utilizes PixArt to generate images based on text prompts. It provides an interactive interface for users to generate, select, and refine images through multiple iterations.

Table of Contents

  1. Setup
  2. Running the Project
  3. Project Structure
  4. Usage
  5. Configuration
  6. Troubleshooting
  7. Contributing
  8. License

Setup

  1. Create a new conda environment: conda create --name pixart python=3.8
  2. Activate the conda environment: conda activate pixart
  3. Install PyTorch and related packages: conda install -c pytorch pytorch torchvision torchaudio
  4. Install other required packages: pip install -U diffusers transformers accelerate safetensors sentencepiece beautifulsoup4 matplotlib ftfy pillow requests scipy tqdm

Running the Project

After setting up the environment and installing all required packages, run the project using: python main.py

This will start the image generation process using the initial prompt defined in config.py.

Project Structure

  • main.py: The entry point of the application.
  • config.py: Contains configuration settings for the project.
  • image_generation_loop.py: Implements the main image generation loop.
  • generate_image.py: Handles the actual image generation using PixArt.
  • display_image.py: Manages image display and user selection.
  • user_input_handler.py: Handles user input during the generation process.

Usage

  1. The program will start by generating a set of images based on the initial prompt.

  2. Images will be displayed, and you'll be prompted to select your favorites.

  3. You can then choose to:

    • Regenerate images
    • Reselect from the previous set
    • Modify the prompt
    • Adjust the temperature
    • Continue to the next iteration
    • Restart the process
    • Stop the program

  4. The process continues until you decide to stop or all iterations are completed.

  5. Selected images are saved in the generated_images folder.

Configuration

You can modify various settings in the config.py file:

  • RESOLUTIONS: List of image resolutions for each iteration
  • NUM_IMAGES_LIST: Number of images to generate in each iteration
  • INFERENCE_STEPS_LIST: Number of inference steps for each iteration
  • DEFAULT_TEMPERATURE: Default temperature for image generation
  • INITIAL_PROMPT: The starting prompt for image generation

Troubleshooting

  • Ensure you have sufficient GPU memory if you're using CUDA for faster processing.
  • If you encounter CUDA out of memory errors, try reducing the number of images generated or the resolution.
  • For any persistent issues, check the error logs and consult the PixArt and Diffusers documentation.

Contributing

Contributions to this project are welcome. Please follow these steps:

  1. Fork the repository
  2. Create a new branch (git checkout -b feature-branch)
  3. Make your changes and commit them (git commit -am 'Add some feature')
  4. Push to the branch (git push origin feature-branch)
  5. Create a new Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

For any additional questions or support, please open an issue in the GitHub repository.