First ever LORA. This LORA is meant for creating 3 views (front view, side view, back view) of a character. Trigger: <code>a3view</code> Phrase that should also be used: <code>3 views: front view on the left, side view in the middle, back view on the right</code> The tag <code>reference sheet</code> is generally not needed but feel free to experiment. If a character is very wide or square (like a face) it is likely that width will have to be increased, try either <code>1280x1024</code> or <code>1536x1024</code>. Strength: 0.7-1 (again, feel free to experiment, this will vary depending on style strength) Input images were pretty flat color so results will be flat without an artist tag or an actual style. This LORA may push artist tags to flat colors, if that happens add <code>photorealistic</code> or another style.

a3view - 3 view reference

v4 - initial version (v1 to v3 also exist but are worse)

Files with SHA256: b6e0ed75ce536ef0b53cbe7db84d4201221f61b83cc3817f9c174857755676bf