You are the blessed one. ...

Making a LoRA with your own dick? Ok. But then putting it on CivitAI for everyone to use? That..... takes some balls.

Hi there, I've had to remove some images from your model. Please feel free to upload age-appropriate images if you want to include people in your display model.

you’ve got some balls 😂

@esc3238 I didn't notice before that you trained this on your own penis, and might I say that the AI version of it is quite fetching. I have mixed feelings about such a homogeneous model, though.

On one hand, it's probably going to fail or at least struggle with darker skin, or differently shaped heads, or circumcised penises, or other differences from your own. On the other hand, by reducing the variation (i.e., the number, types, shapes, and colors of penises), you've increased the stability. So, for anyone who wants a penis just like yours (I'd love to borrow it 😆), your lora will be boss.

Any plans to make it more homogeneous in the future by including other penises? Or maybe an alternate version?

Request: Train it to be "hands free" - no fingers or hands anywhere in the dataset.

Also: Train it on regular SD1.5 so that it's more checkpoint agnostic.

I know it takes many hours to train a LoRA and to curate a new dataset, because I have trained several myself, but... If you want to make the most useful penis LoRA anywhere to be found... Please?

Is this your penis? I'm just gonna call you daddy from now on.

This Penis was really beautiful im getting wet just by looking at it 😅

The penis is the best, it works well with almost all base model, but it seems that the lora will make texture of face worse, for example used with chilloutmix, it would be perfect if it could be fixed.

When trying to switch to the option to use this, it gives an error:

safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooSmall

This popped up as new? What is new? Did you delete the new upload?