I really hope this gets traction and starts getting used more. The facials are incredible. I'd try myself but it costs so much for video here and I always get messed up results.

when you say your lora was trained on videos of a resolution of 211 you mean 211 x 211 square video clips?

Hi

Thanks for your Lora!!.

Do you used a different datasets for High and Low noise ??

For example High noise - video clips
low noise - high quality photos ?

your wan 2.1 version was my favorite but I find the results in 2.2 to be slightly underwhelming, are you using a specific strength or combination with any of the speed loras that present better results in 2.2?

Hello Teacher, why does my video show "Restricted Adult Content"?

If your input image lacks valid metadata, the output will be restricted to PG or PG-13. What do I need to do to see my video?

Don't work for me. Whatever my settings. The penis and, man, man's hand, is always blurry. As there is a DOF focus on the girl. But no DOF, other lora work perfectely. Like the aesthetic cumshot, the penis is not blurry and fully visible

If you're using this for I2V, here's what I've learned. You have to weight your prompts hard. Seriously, weights that will give you nightmare mutants in t2v. Two to three in some cases, if you're not a cunning linguist. And be very, very specific. Exact viscosity, color, turbidity, velocity, the direction and the source, along with all the physics. Keep all your absurdly graphic novelettes in a text box for future use to avoid carpal-tunnel. And pay attention to the notes in the model card. Shift matters a lot when it comes to energetic fluids. That's my 2.9 cents. Worked for me.

proper str for both noises? im getting slightly blurry results

everything works badly! through the ass there is not enough power or even 0 influence(

instructions not clear for what I'm supposed to be doing for i2v

This is really damn good in 2.2 T2V. 2.2's T2V is so good that I don't even feel like I need to use image generation first. Also protip this combines very well with the 2.2 oral insertion lora for better dongs (I have been using 0.5 strength for that and 1 strength for this)

I keep getting locked out from selecting this for generation resources. The issue isn't as bad as it used to be, but this one always seem to wind up on the short end of the stick, I can either cite you and lock out the other seven, or vice vera, so I just go with including as much as I can, but I don't like using stuff without citing, regardless of brow altitude. I use it a lot, and it's very versatile. And so many people comment about how they can't get it to work... anything I use it in is definitive proof that they suck, not you. The site even blocks the metadata I put back in specifically to try to get around the problem. Oh well. Sorry.

none of those cumshot loras work for me. i can somewhat achieve decent results using multiple loras (3 AT LEAST), but it comes complicated to tune up the settings for an optimal result and once the sweet spot is found, it doesn't work into other workflows even copying/pasting with the same settings...

protip if you're using it for anime: add descriptions like milky white or else it comes out transparent

Now that I finally got one your workflows to work on my system, I see that you are genius. Thank you!

In my own experience, all those cum loras give excellent results when used together (2 loras or more) but you have to play a lot with strength and CFG. Fluids are specially complex for the model to depict and very sensitive to micro adjustments. Lightx2v and other acceleration loras strength also affect the result. Remember, ALL lora strength and CFG are the key, high pass for dynamics, low pass for appearance. Once you understand the role of each of these points, you take control.