This is my first lora! Please be gentle :'3
Please post your results if you make images with this lora!
I will tip buzz to my favorites!
I will do my best to answer any questions but first check here-
It's got a lot of moving parts, so this will be a long description, if you'd rather figure it out yourself, I've put all the important prompts in
code bracketsIt's a "Strapped down to a hospital bed, table, or gurney" lora, think insane asylum or feisty hospital patient. Please note the 'straitjacket' mode is the most consistent, if you want arms just strapped to the sides, or especially above the head, be prepared for some inpainting- that being said with this lora + Pony model magic just painting some black blobs on the body parts you want to see straps on will yield some impressive results quickly!
Anyway on to how to use it-
Requires these-
strappeddown, bed strapsstrappeddown is the invoke phrase. Some important helper tags would also be
bondage, restrained, bound (body part)From there you have options/can get creative. There are two main types-
horizontally restrained and
vertically restrainedBoth work but horizontal is more consistent and less gacha dependent because of the training data being mostly that, but vertical can work. This is to determine if your "patient" is upright or lying down, so helper tags are useful here, for horizontally restrained please use
lying, on back, or lying, on stomach, ass,Vertical you will want
upright, and/or upright gurneyTrained surfaces include
gurney, hospital bed, bed, though chairs kind of work too. I haven't tried tables or other stuff yet but you're welcome to give it a shot of course, post results if you do!
If you use gurney I recommend adding
wheelsFinally the restraints themselves, they come in several forms, I'll just list them, go nuts.
straitjacket, bound wrists, bound thighs, bound legs, bound ankles, bound arms, bound neck, bound chest, collar, arms crossed, legs spread, arms up,Take your pick, but you might wanna add some in the negative if you're having issues, like crossed arms if you just want the arms up or arms up if you want them crossed or at the sides.
Oh and also upright likes to add a second person/a dom for your sub, just use solo in positive and multiple subjects in negatives if you want to avoid that.
Enjoy! And lemme know what you think! <3
Description
FAQ
Comments (10)
Been trying to use this lora, but I can't get the eyes to look right at all so far on any testing. Once I figure that out I'll post my stuff
If you are on A1111 or ForgeUI, are you using adetailer? If not I highly recommend you do so, it's pretty much a must for any output that has 'distant' faces/subjects.
https://github.com/Bing-su/adetailer
what's your recommended resolution (any HI-res) for real life pictures using say SDXL or SDXL Turbo ?
Sorry I don't have much experience with either of those so I can't give a super solid answer- My XL experience is primarily limited to pony based models.
I can tell you with a high degree of certainty from my limited non-pony XL checkpoint experience it uses the same resolutions as pony based models- 1024x1024, 1152x896, etc. I guess for an upscaler you can use 4xUltrasharpV10 at 1.3-1.7 increase in size for a photo real model.
However, please note from what I understand pony based loras such as this one don't necessarily work very well with 'normal'/non pony XL models, and the dataset used for this model used only illustrated images so even if it does work with non-pony XL checkpoints its definitely going to make your outputs a lot less photo real.
Edit- sorry misread your question initially, fixed.
If this IS your first lora, I really want to see more from you after getting used to it :D
(seriously, give me more!)
Haha, thank you! It is indeed my first lora, it did take about 7 tries in training to get right though in all fairness. I will make more for sure! I've made more since this and my other one here on civit but nothing worth releasing publicly, I have a bunch of ideas and training data sets I wanna use but it's just a matter of finding the time to toy around.
I am getting the most awesome results with this. Amazing work with the training!
I REALLY NEED a Illustrious version of this I do not like working with Pony
Hi there! Illustrious actually does this concept okay-ish on it's own and there's a good segufix lora that does much the same thing. That being said I probably will get around to making an Illustrious version when I can, since I made this I've switched day jobs and have a lot less time for lora creation. Sorry about the wait!
@Straitjacket So...will I be able to see it this year?I really need it to help me create Illustrious models :(
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.










