This is a pair of LoRAs trained on the WAN 2.2 I2V 14B low and high noise models using 32 81-frame clips. Like my other insertion LoRAs, these models are trained entirely on clips where the man is not initially in frame.

Trigger Phrase:

A man appears and she sucks his penis

I recommend using a CFG of ~3.5 for high-noise sampling and 5-7 for low noise sampling. The examples in the showcase were generated using a constant CFG of 3.5 and 7 for the high and low noise sampling steps, respectively, 24 total sampling steps, a split step of 3, and a shift of 5 for both samplers..

To take advantage of this model's full potential I recommend not using LightX2V LoRAs because they appear to diminish the quality of the results.

CivitAI doesn't allow the attachment of two safetensors files within a single model version, so I attached both model LoRAs within a single .zip file.

If you like this model and want to support the creation of future models, please support my work by leaving positive reviews or posting your generations to the model page.