1. Edit description.csv to specify what prompts to generate images for and how many requests to send (4 images per request), set num_generated column to 0 at first.
  2. Change the API in generate_textures.py to your own, the API website is https://dashboard.goapi.ai
  3. Run generate_textures.py, this sends all requests for generation.
  4. Wait for image generation to complete, wait up to a couple hours (otherwise the tasks expire in Midjourney), then run upscale_textures.py, which upscales each of the four images per request.
  5. After those upscales complete, run fetch_results.py to download the upscaled images.