This model was trained to generate emojis! By nature, the outputs are simplified but are meant to capture the essence of their subjects. The model is good at depicting people and animals. Abstract faces (e.g. smiley/sad faces) aren’t consistent at the moment, but can likely be fixed in future versions.  Join our <a target="_blank" rel="ugc" href="https://discord.gg/BuEEtKnfUK">Discord</a> for any questions or feedback!Model is also available via <a target="_blank" rel="ugc" href="https://huggingface.co/Mobius-labs/emoji_model_msfluentui">Huggingface</a>. Prompting Use "msfluentui style" to trigger the model's learned style "A pirate emoji, msfluentui style" The example images were all generated with very simple prompts, but feel free to experiment with more complex prompts as well as negative prompts. Settings Euler A with 11 CFG and 125 inference steps Credits Thanks to the <a target="_blank" rel="ugc" href="https://pixelpoint.io/blog/ms-fluent-emoji-style-fine-tune-on-stable-diffusion/">creator</a> of the "2D to 3D emoji model" - this larger model extends their work by training on a larger dataset.

Files with SHA256: b717b74f16cfe881a65ea9154e756293ea73c07aee9c81512a9dd46763a86827