Introducing Sigrid for LTX 2.3, my first character LoRA for LTX.

Important: This model is 100% fictitious and is not based on images of any real person. She originated with images generated from still image checkpoints and has evolved from there using only AI rendering. No images (neither actual nor AI-generated) of any real person was ever involved in the making of this model.

This model is based on the BustySigrid models available for Wan, Pony, and Flux. Sigrid now has her own voice and personality to go with the visuals you all are used to. Sigrid renders well in both Image-to-Video and Text-to-Video scenes.

• Text-to-Video: recommend 0.8 strength

• Image-to-Video: recommend 1.0 strength

All of the sample videos in the showcase were rendered at those values. You can usually use strength 1.0 for text-to-video, but you have a higher chance of not-quite-right anatomy, especially with certain schedulers, so it's safer to use 0.75 or 0.8.

No topless or nude samples were included in this dataset because I haven't solved the riddle of teaching LTX how to do that without making truly gruesome anatomical decisions. If anyone out there has had success at doing this, I'd love to hear your advice on sample images/movies, captioning, and training settings. In the meantime, Sigrid is keeping her clothes on unless you bring in help from another LoRA. For those of you hoping to direct some more explicit LTX scenes with Sigrid, sorry about that.

For now, I hope you enjoy making videos (with sound!) about Sigrid. Please don't forget to tag this model when you post videos so I can see what Sigrid is getting up to when I'm not around.