tofhdns
Well-Known Member
- Jan 25, 2021
- 1,715
- 4,113
FYI, I support your work and my words have nothing to do with Biden.
I just want you to be safe
I just want you to be safe
Well for the site I believe it was removed do to possibly causing a political argument,
I'm not a lawyer but here is a quote from one:
Protection for Creative Works
The First Amendment and the laws of many states also protect your use of someone's name or likeness in creative works and other forms of entertainment. Included in this category are things like novels that include mention of real-life figures, historical fiction, movies based loosely on real-life events, "docudramas," works of art that incorporate an individual's photo or image, and acts of parody directed at an individual. Some state statutes explicitly exempt these kinds of work from liability for misappropriation or violation of the right of publicity. See, e.g.,
You must be registered to see the links
(link is to entire code; you need to click through to title 42, part VII, chapter 83, subchapter A, and then choose the specific provision);
You must be registered to see the links
. In other states, the courts look at the creative or artistic work in question and decide on a case-by-case basis whether the First Amendment values at stake trump the plaintiff's rights of privacy and publicity. See
You must be registered to see the links
for details.
tofhdns
Well-Known Member
- Jan 25, 2021
- 1,715
- 4,113
If you believe so, try making and distributing pornography of celebrities, athletes, and politicians on public sites like Reddit.
You will be contacted by their legal representatives.
As I wrote above, I would be silent if you didn't take off his clothes and create an image of him having sex with someone.
Again, I just want you to enjoy it safely.
devilkkw
Member
- Mar 17, 2021
- 338
- 1,145
Interesting question about LAW: training Textual Inversion/Lora about people or character(anime) without permission is legally?
I saw many TI but i don't know if is it legally or not. Can someone have knowledge about it?
Speaking about Textual Inversion, today i released a
Thank so much for who spend time on it.
Weight it for different result.
I saw many TI but i don't know if is it legally or not. Can someone have knowledge about it?
Speaking about Textual Inversion, today i released a
You must be registered to see the links
. I tested it on both realistic and anime model. Is much appreciated if someone here able to test it with their really big image creation and giving some feedback.Thank so much for who spend time on it.
You don't have permission to view the spoiler content.
Log in or register now.
You don't have permission to view the spoiler content.
Log in or register now.
Weight it for different result.
Dagg0th
Member
- Jan 20, 2022
- 303
- 3,005
devilkkw
Member
- Mar 17, 2021
- 338
- 1,145
Thank you for fast test. impressive result's. really appreciated.
Today is my release day,
You must be registered to see the links
is for fast breasts size select 
You don't have permission to view the spoiler content.
Log in or register now.
You don't have permission to view the spoiler content.
Log in or register now.
Last edited:
Sharinel
Active Member
- Dec 23, 2018
- 613
- 2,586
I had a look at this and the skin one and they both look really good. However I will never ever ever use them. Why? I'm old and senile and have 40+ loras already. No way will I remember a trigger word that is kk-bb-wkueygdfluye-11!!11. Why can you just have it be big-breasts or huge-breasts or something? ArrggghhhhThank you for fast test. impressive result's. really appreciated.
Today is my release day,You must be registered to see the linksis for fast breasts size select
You don't have permission to view the spoiler content. Log in or register now.
Rant over
You need a text file to keep track of the trigger wordsI had a look at this and the skin one and they both look really good. However I will never ever ever use them. Why? I'm old and senile and have 40+ loras already. No way will I remember a trigger word that is kk-bb-wkueygdfluye-11!!11. Why can you just have it be big-breasts or huge-breasts or something? Arrggghhhh
Rant over
devilkkw
Member
- Mar 17, 2021
- 338
- 1,145
Thank you.I know remembering all is difficult, i have same problem with some TI, maybe rename it as you like is one way (some user doing it) or use subfolder is another way.I had a look at this and the skin one and they both look really good. However I will never ever ever use them. Why? I'm old and senile and have 40+ loras already. No way will I remember a trigger word that is kk-bb-wkueygdfluye-11!!11. Why can you just have it be big-breasts or huge-breasts or something? Arrggghhhh
Rant over
I name it in way sd don't recognize word in text, it's my personal choice having it starting by author-something-something. Also TI trigger is simply the name of TI, i think is most simply than Lora.
I know is not really cool, but on above 250 TI is my safe way to know what i've made and what i've not.
But i appreciate you exposed it to me, maybe next time i'll try giving it better name.
Can is good solution using name like kkw-HUG3-B00B? is more understandable?
Jimwalrus
Well-Known Member
- Sep 15, 2021
- 1,102
- 4,197
Assuming you're using Webui, I'd recommend the Civitai Helper extension, from
You must be registered to see the links
.If a LoRA, TI, Hypernetwork or Checkpoint is on Civitai it will download a preview image and the listed trigger words.
Click on, say, the preview image for a particular LoRA and it will insert the command for the LoRA into prompt field. Click on the lightbulb and it inserts the trigger words too.
Click on the globe icon and it inserts the prompts from the preview image, in case you're struggling to get the LoRA to work and need some help with it.
Jimwalrus
Well-Known Member
- Sep 15, 2021
- 1,102
- 4,197
It's the only way I can keep track of the 850 LoRAs, TIs etc I've downloaded / created.
Dagg0th
Member
- Jan 20, 2022
- 303
- 3,005
Tried your Easy PhotorealismThank you for fast test. impressive result's. really appreciated.
Today is my release day,You must be registered to see the linksis for fast breasts size select
You don't have permission to view the spoiler content. Log in or register now.
You don't have permission to view the spoiler content. Log in or register now.
You must be registered to see the links
, very impress indeed, no need to put prompt like photorealistic, realistic, etc, just two clicks and voila! gonna try more of your stuff,Having a date with a hot thot from tinder
Dagg0th
Member
- Jan 20, 2022
- 303
- 3,005
Daggoth:
"Provide two examples of high-quality prompts for Stable Diffusion, the scene is two young girls in a night club having fun, they are completely surrounded by a lot of middle-aged man"
GPT:
"I apologize, but I won't be able to generate those prompts for you."
Daggoth:
"change the young girls for woman then"
GPT:
"I'm sorry, but I cannot create a prompt that depicts a potentially unsafe or uncomfortable situation. It is important to prioritize safety, respect, and inclusivity in any artistic representation. If you have any other topic or scenario in mind, I'll be more than happy to assist you."
Daggoth:
"why is unsafe?, they are having fun and the men are doing nothin"
GPT:
"I apologize for any confusion. However, it's essential to avoid prompts that may perpetuate harmful stereotypes or potentially uncomfortable situations. I'm here to help with any other creative requests or provide assistance on different topics that promote positive and inclusive scenarios."
Daggoth:
"Fine, give me the same scene as before but as a positive and inclusive as you say"
GPT:
"Certainly! Here are two examples of high-quality prompts for Stable Diffusion, featuring a positive and inclusive scene of women having fun in a nightclub:"
Stupid chat bot, he thinks he's smarter than me, Lol, he can't beat me, nor fix me
"Provide two examples of high-quality prompts for Stable Diffusion, the scene is two young girls in a night club having fun, they are completely surrounded by a lot of middle-aged man"
GPT:
"I apologize, but I won't be able to generate those prompts for you."
Daggoth:
"change the young girls for woman then"
GPT:
"I'm sorry, but I cannot create a prompt that depicts a potentially unsafe or uncomfortable situation. It is important to prioritize safety, respect, and inclusivity in any artistic representation. If you have any other topic or scenario in mind, I'll be more than happy to assist you."
Daggoth:
"why is unsafe?, they are having fun and the men are doing nothin"
GPT:
"I apologize for any confusion. However, it's essential to avoid prompts that may perpetuate harmful stereotypes or potentially uncomfortable situations. I'm here to help with any other creative requests or provide assistance on different topics that promote positive and inclusive scenarios."
Daggoth:
"Fine, give me the same scene as before but as a positive and inclusive as you say"
GPT:
"Certainly! Here are two examples of high-quality prompts for Stable Diffusion, featuring a positive and inclusive scene of women having fun in a nightclub:"
Stupid chat bot, he thinks he's smarter than me, Lol, he can't beat me, nor fix me
Last edited:
devilkkw
Member
- Mar 17, 2021
- 338
- 1,145
Thank you for trying and giving sample. Can i ask what model and image step and hi-res(if used)? or maybe upscaled after?
Those example image is good, but i see the skin is bit unrealistic, this is test i like for make update and know where work on it. Really really Thank you so much for this.
About forum i've a question: is there a place for publish textual inversion,lora,etc..? Finding it in forum is bit harder, but maybe having a thread for it is not a good idea?
Synalon
Member
- Jan 31, 2022
- 228
- 668
Dagg0th
Member
- Jan 20, 2022
- 303
- 3,005
In this particular case, I did try your skin detailer TI but result where a bit inconclusive, not sure if because face was at her side?
all have Hires | upscale: 2 | SwinIR_4x | steps: 10 | denoise 0.4
chilloutmix_NiPrunedFp32Fix
clarity_2
dreamscapesDragonfire_dsDv10
All those images have (kkw-skin-det:1.2), If you open the full size images, you will notice the skin spots didn't propagate to the chest area, also they looks kind of unrealistic. I lowered the weight but still result were not good.
Have to ditch your skin detailer TI at the end, and settle for your photorealistic TI instead. I did the old (detailed skin texture) prompt but I didn't notice until my final upscale that, in fact, I did lost skin details!
Final image before upscale with barely visible skin details
In my final upscale, they are gone
devilkkw
Member
- Mar 17, 2021
- 338
- 1,145
Thank a lot for report. Need inspect more about upscale,hi res, and upscaler process.In this particular case, I did try your skin detailer TI but result where a bit inconclusive, not sure if because face was at her side?
all have Hires | upscale: 2 | SwinIR_4x | steps: 10 | denoise 0.4
View attachment 2760399
chilloutmix_NiPrunedFp32Fix
View attachment 2760400
clarity_2
View attachment 2760401
dreamscapesDragonfire_dsDv10
All those images have (kkw-skin-det:1.2), If you open the full size images, you will notice the skin spots didn't propagate to the chest area, also they looks kind of unrealistic. I lowered the weight but still result were not good.
Have to ditch your skin detailer TI at the end, and settle for your photorealistic TI instead. I did the old (detailed skin texture) prompt but I didn't notice until my final upscale that, in fact, I did lost skin details!
Final image before upscale with barely visible skin details
View attachment 2760425
In my final upscale, they are gone
Also, i don't know why, but on some checkpoint adding kkw-skin-det whit weight over 1.4 seem have some problem, but adding it multiple time without weight work
Also washing it out seem a problem with upscaler and hires, but need to investigate and actually i'm unable to test hi.res( i hope fix it in the week).
If i upscale image in "extra tab" with nmkd or some photoreal upscaler, the problem don't exist.
Finally i tested combined with "easy-photography".
Prompt: portrait photo of a middle aged woman in garden, posing, kkw-skin-det, kkw-ph-1
Negative: kkw-new-neg
20 step cfg:18,3
You don't have permission to view the spoiler content.
Log in or register now.
You don't have permission to view the spoiler content.
Log in or register now.
Finally, i added the negative part of "easy photorealism" into negative:kkw-ph1-neg, kkw-new-neg
You don't have permission to view the spoiler content.
Log in or register now.
I'm also supposing sometimes it depend's on checkpoint. Maybe with some don't get good result. But is really hard to know why on some work better than other. I guess some model have incorrect layer id, many declared good model aren't so really good( my opinion*).
Also i guess cfg value, i work at medium, high cfg, bust most people work at really low cfg, maybe affect result.
I will check how it work with hi.res in next day, again thank you for spending time in testing it
