Lora for doing chipmunking poses. You have 3 versions to choose from.
If you want to support me head to: https://ko-fi.com/seaeng
Description
Here we go again, version 2.0
Trained with over 500 pictures!
It's very close to 1.0 but more honed and doesn't force towards realism as much.
The main trigger word is chpmng, since the former "chipmunking" could trigger unwanted chipmunk / chipmunk features.
The pictures used for trained is more varied with positions, should give better results on different positions.
I've recreated the pictures used in V.1 showcase with the same prompt so it's easier to see what differs from the versions.
Eager to see what you guys will generate 😄👍
Ps. If you want to make a tip but dont have buzz to spend, post one picture with only this Lora as resource, then I receive buzz.
FAQ
Comments (8)
u deserve some tip my friend
If you want to make a tip but dont have buzz to spend on tip, post one picture with only this Lora as resource, then I receive buzz 😄
Hi, this is amazing! I love this! I noticed that this Lora applied a great 2.5D realistic style to Pony and even adds stylized skin and lighting realism to Everclear. How did you make this? What was your training data like? Could we get a general style Lora to apply this 2.5D look to all of our images? I see huge potential beyond just individual sex poses here. Albeit this pose is one of the best I've seen yet. Would love to help out, but I don't have hardware, I just use Civit and have yet to make my own models. Would also love an article on how to replicate this. ♥️♥️♥️🐿️
Thank you 😄❤️ Yes it turned out to be a general style Lora without me intending to. Sounds really interesting to create a specifik style Lora.
I've used civitai:s Lora trainer and this is the first Lora I created. I followed this guide when preparing and gathering training data,
https://civitai.com/articles/4100/how-to-create-a-style-lycoris-for-ponyxl-on-low-vram-locally
@SeaEng so you used that guide for training on Civit? It looks like it's tailored for Kohya and Automatic1111. So I presume you trained on Pony. Were you images photographs or 3D? About how many images? Also, if you trained on Civit, did it cost a lot of Buzz?
@SeaEng BTW, I found a pretty good combo of LORAs and tokens that works for my liking for very photorealistic Pony Diffusion thanks to your LORA. Great stuff man! Thanks again for making this.
@SeaEng I know I keep bugging you, but I've started using this LORA for more generalized sex scenes/poses and all the blowjobs have the head turn and cheek bulge, could you maybe prioritize a generalized LORA? I will help however I can if you'd like. I have tried to replace this with Photo Pony, RELSM, Photorealistic Pony, non of them work the same and all look 2.5D or less. Yours is the closest to realism.
This is great. Are your training images real photos or animation. I like the skin and lighting detail this lora adds. Also, what is the quality of the images? Are they like professional studio shots or amateur porn? About how many did you use? Any info is helpful as I want to make Loras like this myself. Thanks!
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.