NOTE: I am taking a break from Civitai, so I will not be responding to questions.
Some people have requested my workflow so I am posting it here. It includes prompts for all the videos I've generated.
This is a complicated workflow for advanced users. It allows for generating videos in batches. It can detect the type of input image based on specific filenames e.g. "missionary" and then apply one or more prompts to that image. If there are multiple prompts that match, it will generate multiple videos (e.g. "preblowjob" image --> "blowjob" video + "facial cumshot" video).
This uses Lightx2v, GGUF, and SageAttention and generates videos in less than 2 minutes on my 3090. It also includes an upscaler which uses the 1.3B WAN model, as well as a 60fps interpolation.
I will not include instructions on how to use this, but here is post to help you install SageAttention: