<h3 id="description">Description</h3><p>This collection of LoRAs generate some fictional women. These are trained from a single AI generated image that I wanted to replicate more of it, so they will not trigger in all contexts. These LoRAs are trained using Twisted-Reality-XL, so I am not sure what is their effect on other base models.</p><p>Many regularisation images were used for this, about 10x the number of steps on the actual reference image. This gives results more aligned on the base dataset outputs, while not locking the base model.</p><p></p><h3 id="trigger-words">Trigger words</h3><ul><li><p>v1 : bl0nd3 woman</p></li></ul><p></p>

Files with SHA256: 0a0f9acea4ca53b90870b74a1d6d964f526ae1719006cfe603a0050f6a8e8e4e