Skip to content

Latest commit

 

History

History
79 lines (60 loc) · 4.67 KB

README.md

File metadata and controls

79 lines (60 loc) · 4.67 KB
Raw

brAInstormer

brAInstormer is a robust and multifaceted creative suite that leverages the power of GPT, Stable Diffusion, Dalle2, and other forms of AI in order to provide a diverse set of functionalities from a single screen.

The core script features a simple but powerful GUI that enables capabilities such as text-to-image generation, text-to-animation generation, experimental file conversions, a notepad, ChatGPT, and much more, with everything fitting on one screen.

It is also packed with various randomization features powered by GPT, making it a one-stop tool to inspire your creativity and streamline concept design.

Main Features

  • Dalle-2 AND Stable Diffusion text-to-image generation.
  • Edit, filter, and add font to images with 30+ visual effects that include pixelate, solarize, vignette, greyscale, sepia, mirror, flip, VHS glitch, and many more.
  • Upscale Images: Enhance the quality of any imported image and enlarge it to 2048x2048.
  • Image-to-Image generation: Import any image then transform it using text prompts to guide the generation process.
  • Image variations: Create endless versions of your generations with the click of a button.
  • Text-to-Animation: Generate animations using text prompts, fully equipped with camera and frame controls, 10+ style presets, randomization tools powered by GPT, and more.
  • Image-to-Animation: Import any image and use it as a starting prompt to shape 48-frame animations.
  • ChatGPT-4: Host up to 3 simultaneous conversations with ChatGPT. Name and assign them different roles and bounce between each chat seamlessly. All conversations can be quickly saved, cleared and imported at any point.
  • Experimental file conversion: Convert images to audio, MIDI, and more.
  • Notepad: A convenient tool for note-taking, copy-pasting, or anything else you might need it for. The notebook is also fully equipped with 10+ GPT-powered concept shufflers that can randomly provide movie scene suggestions, art styles, trinket lists, and more.
  • Sketchpad: A simple sketchpad for concept doodles with full color spectrum control. Sketches can be saved and cleared at any point.
  • Multiple GPT-powered randomization algorithms to eliminate your creative block.
  • All of the above on a single screen means no more tab jumping or app switching. In fact, you don't even need to scroll.

Screenshots

Software Screenshot 1

Software Screenshot 2

Set Up

  1. Obtain Stability AI and OpenAI API keys. Register on their websites, check their pricing details, and get your own personal API keys:

  2. Install Python 3: https://www.python.org/downloads/

  3. Download brAInstormer.py, words.py, and video.mp4 from this repository

  4. Create a single folder and put all the downloaded files in it. All generations will be saved in this folder.

  5. Install necessary 3rd party modules:

    pip3 install opencv-python
pip3 install stability-sdk
pip3 install "stability-sdk[anim]"
pip3 install tqdm
pip3 install openai
pip3 install pydub
pip3 install midiutil

  6. Download and install ffmpeg: https://ffmpeg.org

  7. Open brainstormer.py with your Python script editor. Replace 'YOUR API KEY GOES HERE' with your personal API keys in the appropriate places of the script.

  8. Run the module and enjoy brainstorming!

Important

  • All generations auto-save as the name of the initial prompt in the root folder.
  • I'm brand new to coding. I began self-teaching 6 months ago with the help of ChatGPT in order to build this. So this whole script is the work of a freshman. For this reason, all feedback is absolutely welcomed and encouraged. Just please be kind :) I'd like to make this as awesome as possible.
  • This has been tested on four computers. It seems to work correctly with the latest Mac OS. It has been tested on Linux, but the display is a bit wonky. It has not yet been tested elsewhere.
  • I encourage you to explore the script, customize it, and make it your own!
  • For Stable Diffusion's Text-to-Image, you must assign a step amount before clicking on 'generate image'.
  • For generating animation, fill out all prompts and select a preset. The starting prompt must be at Frame 0.
  • Animations take time. To expedite the process, reduce the seed and the total frame count (length).
  • Image-to-Animation generates only 48 frames and is rather wonky.

Support

If you found brAInstormer helpful and would like to support its development, consider buying me a coffee:

Support via PayPal

License

This project is licensed under the MIT License