Sandpit environment for experimenting with different stable diffusion models and prompts.
The two models used in this sandpit are the Stable Diffusion v1.5 from Runwayml and OpenJourney from Prompthero.
I have developed a function that allows you to process multiple images from a single prompt, these images are then stored in a list allowing you to select the one which you like the best.
The default is 2 images, it takes ~3-4 minutes to process a single image.
Using the prompt:
temple in ruines, forest, stairs, columns, cinematic, detailed, atmospheric, epic, concept art, Matte painting, background, mist, photo-realistic, concept art, volumetric light, cinematic epic + rule of thirds octane render, 8k, corona render, movie concept art, octane render, cinematic, trending on artstation, movie concept art, cinematic composition , ultra-detailed, realistic, hyper-realistic, volumetric lighting, 8k –ar 2:3 –test –uplight
The OpenJourney model produced the following images:
The Stable Diffusion v1.5 model produced the following images:
