My Gratitude (Pixiv Fanbox)

Well, it's about time we gonna end this soon...

I wanna say that I really - really appreciate all of your support up until now, I don't think I'll be where I am right now without your support, so thank you 🙇‍♂️

A little (actually long) story telling from me now, I remember back then when I first found out about Stable Diffusion on like October 2022, I was like "surprised" there's an AI that can make a great anime picture by just prompting from our own imagination... and then my fricking perverted mind start to think... can this make a girl in damsel situation? and when I tried it, holy molly it can, although it is not as easy as prompting a "normal" anime picture.

From that, I start to mess things up with it, playing with the prompt, and start to post it online too on my deviantart, I got quite a backlash back then someone screenshooting my profil and make a warning that "this guy make a DID pictures with AI" some people believe it, some people don't and saying like "there's no way and AI make that", I don't care about that at all and the controversy that's still sorrounding AI image generation up until now, I just want to make the best picture possible.

While I'm happy making pictures by prompting, then the development of AI image generation start to kick in, and one of it is "textual inversion", a technique for capturing novel concepts from a small number of example images in a way that can later be used to control text-to-image pipelines (from hugging face), TLDR, it's a technique to tell the AI on how to draw something "new" for them, I start to think again like : "hmm, if I can make the AI know how a tape gag looks like, how the rope should go around the body and such, would it make it easier for me to make a great DID picture?" and thus, I try to exploring that part of development, which ending in, I can't train with my old PC spec, which is i5 4th gen with GTX 1650 Super 4GB, and with that limitation, I start to think to make this Fanbox, a way for me to get a financial support for a better rig to train the AI for a better DID pictures for everyone, with an exchange of me keep learning by sharing this "picture pack" for the supporters who help.

At the early 2023, I finally got plenty enough supporters or FANs, and with the help of my own earnings from my main job too, I finally be able to have my current rig (i5 13th gen with RTX 3060 12GB), and the AI development is also going faster too! so previously we used textual inversion, now we have LoRA, a training technique that is used for fine-tuning Stable Diffusion models, easier to train with a great output results, I start to learn how to train, experimenting stuff and here we are... the DID AI image generation "community" has going so far, lot of LoRA that help to make a better DID pictures has been released by me and others too, we even have a discord server now which already have a member of 250++ and keep going up!

I can't thank you enough for what we've been through so far 🙇‍♂️

But starting tomorrow 25 July 2023, Fanbox is going to slap us on the face with their policy, which I can't do anything at all except by just moving from here because honestly, the financial support from all of you up until now is really - really helping me to keep getting better in making and developing the DID AI image generation that I've been in love to since forever.

This is might be the end "here", but I will keep going, developing and making DID AI pictures for all of you until we might found that one sweet spot, where everyone no matter how much good or bad their skill in drawing, can make a great DID pictures themselves, BE FREE AND EXPLORE YOUR IMAGINATION WITHOUT LIMIT!

So, thank you thank you thank you, I can't stop saying it and see you... on the "other" side.