Wan-2.2-I2Pee: Blink Piss LoRA
Update: OmniDream Private Access
OmniDream-Omega unlocks brand new all-in-one creative power, it is a groundbreaking blink LoRA capable of doing intercourse (multiple positions, anal/vaginal), watersports (male/female, self/partner), cumshot, object insertion, and much more. Available to subscribers on
ko-fi.com/obsxrver today.
Get access to OmniDream for $14.99 by subscribing on ko-fi
Update: I2Pee-4 Public Release
I2Pee-4 is the best freely available blink piss LoRA today. You can download and run it for free now.
No GPU? No Problem
Run ComfyUI on a cloud GPU using Vast.AI with
this templateWORKFLOW:
Download the
workflow file and drag into ComfyUI.
SageAttention:
The Vast.AI template already provisions the instance, downloading all necessary LoRA Files and installing SageAttention. However, if running locally, you will need to install SageAttention to run the workflow:
pip install sageattention
Technical Explanation
Blink introduces a novel training methodology for Image-to-Video (I2V) diffusion transformers, termed Identity Invariant Action Mapping (IIAM). This technique addresses the persistent challenge of identity drift and hallucination in generative video by decoupling subject identity from temporal action dynamics.
Traditional I2V training often results in the model creating a smooth, continuous video, Blink teaches the model to perform a hard-cut transition, switching to a completely new scene, while preserving the facial and identity characteristics of the input image. We achieve this by utilizing a specific data preprocessing strategy:
-
Temporal Splicing: Training data consists of a 0.3-second static frame (the "before" image of the subject. Derived from the video using Qwen-Image-Edit) prepended to a 4.7-second action sequence (the "after". Whatever you are training the LoRA to do). This "jumpcut" forces the model to attend to the initial static latents as the ground truth for identity.
-
Identity-Agnostic & Discontinuity-Forcing Captioning:
- Template: "a [man|woman], jumpcut, after the transition, [he|she] is (any change in appearance), [he|she] is (the action happening in the "after" sequence)"
By stripping any uniquely identifying descriptors (e.g., hair color, facial structure, clothing), using only generic identity references ("man", "woman", "he", "she"), the model is forced to map the learned motion vectors and physics simulations directly onto the visual topology provided in the initial frame. The model learns to bind the discontinuous transition - the "blink" - to the "jumpcut, after the transition" tokens.
Future Applications:
The IIAM/Blink framework is content-agnostic and can be applied to any set of action clips to create modular, "plug-and-play" motion LoRAs for generative video workflows.
Examples
Note: All Reference Images are Generated using Z-Image-Turbo and do not feature real humans. These examples were created using Version 3. I have not yet created example videos for Version 4.
| Video | Prompt |
|---|
| a woman, jumpcut, after the transition, she is nude, lying on her back with her legs back holding her ass with both hands, she is pissing on her own face, piss splashes on her face and soaks her hair |
| a woman, jumpcut, after the transition, (she is nude) she is on the ground on her hands and knees, a nude man is standing beside her on the left, he is holding his penis in one hand, with his other hand on her head, he is pissing on her face, she is disgusted, she shakes her head, drenched with piss as it splashes in her eyes and runs down her body |
| a man with a woman, jumpcut, after the transition, he is nude, sitting down, she is nude sitting on his penis, she is grabbing his penis, while they are tongue kissing, pissing her, piss from his penis sprays on her, completely soaking her body |
| a man, jumpcut, after the transition, he is nude, lying on the floor with his legs spread, he is holding his penis, he is pissing on his own face, piss splashes in his eyes and hair, soaking his face |
