V1.5
is in beta at the moment... its trained on only videos for better motion and breast recognition
will post examples as time goes on, But cant train and play :(

still never tried Image to video





10k steps, probably stopping at that

all video examples on 1.0 version include the exact workflow.

this does require other loras sometimes to work..
its not perfect. lower strength for clothes to work better

Examples are just the best version of ideas i can think of.

ADVICE use with LTX2 better NSFW motion - v1.0 | LTXV LoRA | Civitai
@ like 0.4 strengh
fine tune the str's yourself it seems different for everyone depending on other LoRa's

all examples are Text to video, its not trained for I2v (IVE NOT TRIED)
THE WORKFLOW IS I2V, i linked the black starting frames i use. in the workflow. (change resolution accordingly keep the aspect ratio the same as your black frame)