*** Version 2 Info ***

Version 2 includes some videos in the training data, which seems to help with prompts involving motion.

One thing I've noticed is that it really likes giving women long fingernails; I must have missed it in the training data and failed to tag it out. But I've just been putting "long fingernails" in the negative prompt and that seems to fix the issue pretty well.

*** Version 1 Info ***

A Lora for Wan 2.1 to help out with bigger natural breasts.

It's trained on the 14B model.

It works for Image to Video and Text to Video.

For I2V I'm using wan2.1-i2v-14b-480p-Q5_K_S.gguf and for T2V I'm using wan2.1-t2v-14b-Q5_K_S.gguf; it works on these, don't know how other models will perform.

The sample images were at a strength of around 0.85 to 1.00.

There aren't any specific keywords, check out the prompts on the sample images for what I use.

Enjoy!