This model was trained on nearly 10k high-quality public domain digital artworks with the goal of improving output quality across the board. We find the model to be highly flexible in its ability to mix various styles, subjects, and details. We recommend resolutions above 640px in one or both dimensions for best results.
You can try this model and several others for free at
seek.art.
We also recommend an inference tool supporting prompt weighting and high resolution optimization / fixing for best results. We suggest
InvokeAI as a sensibly licensed and fully featured open-source inference tool.
The above example images including the prompts and all relevant settings are available
here.
Additionally, search thousands of high quality prompts on
seek.art for free.