MysticPilgrim's Blogs

Greetings,

I've been a long-time lurker on this site.  I've always enjoyed it.  Now, I want to give back, but it's going to be a long project. 

My goal is to train a Stable Diffusion based model to recreate the human for very accurately.  With such a model, we can recreate fantasies in a way that simply wasn't possible before.   For example, all we would need to do is type "20 year old nerdy teenage boy being pounded by a 30 year old muscle daddy in the middle of time square" and BOOM.  There it is.  At least that's the plan.

Here's were my request for help comes in.

I need high resolution images of people of all ages, genders (in this case male), ethnicity, body types, etc. I need to clothed, unclothed, partially clothed, etc.  Taken from different angles, lighting conditions, and settings.   I also need them both posing and doing different activities.  The only thing we need is that the images are high resolutions (Ideally no less than 800x800 but higher is better as it provides more training data per image) and the human being himself needs to be clear and crisp, as that's what the model will be trying to learn.   Also, for nudes, we would need big cocks, small cocks, bushy pubes, cut, uncut, etc... Basically with every aspect of the person, we would need as much variety as possible. Variety and quality - the better the input the model is trained on the better the model will be.

In exchange for the assistance, I'll be for the GPU and Supercomputer time for training and model manipulation.  I'll also make the model free and open source, posting it on GITHUB and/or huggingface.

Anyway, thank you all for the help.  I hope you have an amazing day.