Not my TI. The original is from https://huggingface.co/Xynon/models/blob/main/experimentals/TI/bad-image-v2-39000.pt
If someone knows the creator please contact him so this can be moved to his account.
I just uploaded it as someone was asking about this embed inside one of my sample images.
Edit: Unarchived it as it seems it was unusable in the civit generator.
NOTE:
Hmm I am starting to think the best solution is to archive this so it is findable but people will have to get the original at hugging face. So I will do so.
Description
FAQ
Comments (13)
Wasn't this posted already?
Don't know. The checksum doesn't detect anything. I asked in discord and no one seems to know.
Basically, just select this LORA on the negative prompt and enter the trigger word "bad-image" and it will take effect?
Basically yes. Just a couple of things, this is a Textual inversion those work based on tags instead of adding stuff to the model like loras do(which is why they are so light.). The other is they don't have trigger per se, what you need to use is the name of the file, in this case using bad-image is liable to do nothing, you must instead either rename the file to bad-image.pt or instead invoke it by typing bad-image-v2-39000-neg(I think, I am not sure what name civitai gave it). My sample images are just tagged "bad-image" because that's it's name in my PC but civitai loves to rename stuff.
Also make sure you put it in stable-diffusion-webui\embeddings
@knxo Thank you for your response.
The file name was "bad-image-v2-39000-neg.pt", so I will change it to "bad-image.pt" and use the command "bad-image" and see what happens.
(The folder to put the files in is not wrong)
People should know this works best with Stable Diffusion 1.4. It's over 8 months old, and not worth installing today despite appearing on some famous prompts on this site.
It seems to actually work well with the NAI models family. I wouldn't recommend it for anything else but from testing it, it does work fine in 1.5 for anime models. When i originally found it I was just digging trough negative embed repos, i used it and i liked the results, so i am not sure if i am the origin of it's usage epidemic, but that is likely the case. I mostly added it as people kept asking how to recreate some of my images and it was easier to simply upload it. It is not as neutral as other negative embeds but i do like the bent to a slightly more sober style it provides.
It works fine for me in all models based on SD 1.5 but if you know of better alternatives, feel free to mention it.
It works fine for me in all models based on SD 1.5 but if you know of better alternatives, feel free to mention it.
There is no better NAI 1.5 negative. I will not tolerate this heresy :P
why cant i downlaod this??
This embed is not mine, but it is useful and i wanted it to be findable so I set it as archived(So as to try to not claim the extra downloads for my account) and put a link to the original repo in hugging face. Just download it there and leave a thanks to the author.
Details
Files
bad-image-v2-39000-neg.pt
Mirrors
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000-neg.pt
bad-image-v2-39000.pt
bad-image-v2.pt
Bad Image v4.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
Bad Image v4.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
Bad Image v4.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
Bad Image v4.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
BadImage.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
Bad Image v4.pt
Bad Image v4.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
NEGS Bad Image v4.pt
bad-image-v2-39000-neg.pt
BadImage.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
BadImage.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2.pt
bad-image-v2-39000.pt
NEGS Bad Image v4.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image.pt
badImageV239000_v10.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image.pt
bad-image.pt
Bad Image v4.pt
Bad Image v4.pt
bad-image.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt
bad-image-v2-39000.pt


