There was. I downloaded them both. Not sure what is going on.

@LFMX Shhould i delete the two epoch 2750 ones i downloaded earlier?

thank you :)

I'm curious too. A lot of times with i2v loras I'm not sure how to do the starting images. For this one there's some other versions like sdxl, maybe that can get it going

If yall are using lightx i2v checkpoints or loras , you can actually us a solid black picture as a starting image and prompt it as you would for 2tv (very descriptive) you can basically use i2v just like t2v ! Loras too! This 100% works, no BS !

@jynkz41 wow, thanks. I got to try this

Exactly. Thats why I2V loras are brainless and stupid. The vast majority of T2V loras work on I2V. This entire base model exists to fuck the community

@jynkz41 Not always.... T2V loras are what this is for. Youre literally using the I2V lora in a brain-damaged state. Creators just shouldnt even make I2V loras. They seem ... useless.

@llhappyll822 well let me tell you when image to video work like a champion,you start with an image of a woman just standing there minding her own business and then it blinks to her performing some useful task with her mouth. When you have that kind of scenario, image to video is priceless. For this one I think it really wants you to be close to that position maybe.

@llhappyll822 Why so hostile, dude? Clearly image to video has huge demand for taking existing art of any kind and then animating that and keep the exact likeness. You say t2v can work, but I wouldn't say it works as well. There's a workaround, yes - just like there's a workaround to use a black pic and prompt what it transitions to for i2v. But for some reason the other way around you describe the use of the workaround "brain-damaged"...

If you're this unhappy and this demanding and can't even afford any respect to anything that isn't going your way, the only way to get exactly what you want is to make it yourself.

actually the black image thing works great!

Time to make your own Max Hardcore. Lol

@LFMX  That would be great.

Yes definitely works. Endless ways to do that, be creative. Even with Wan2.1 it was possible. Many examples here on Civitai to help you prompt. Or use FLF2V: Wan manages intermediary frames impressively well.

Be creative! Prompt what you want to see, or add a black area to your picture and / or use a lora which can draw a penis like ultimate deepthroat or oral insertion and prompt the man to appear.. there are so many possibilities.

That is a possible Prompt (A woman lies on her back at first, no man in the room):

"The woman turns onto her back. A naked man appears (behind her head:2) and puts his penis in her mouth. Then reverse deepthroat, in and out, deep throat fuck." That worked for me. The sentence in the brackets means, that ComfyUI preffered the order. Without that the man appeares on the false side.

are those your str settings for the high and low loras?

@LFMX powerful! will try in multiple steps. ty for your time