TextualInversions

Info about the person:

Roberto Portales is an actor, fitness trainer, fitness model, and captain of the West Point Army.

Note about the model:

This model was requested from here. It was trained on a dataset of 67 images, I always try to add more images if their quality is good, and this model also shows good results in the DeepFace tests.

To request a Textual Inversion:

It's too hot on Civitai, feel free to get me a drink!

Liv A is 34 year old Russian nude model

Here goes the Iron Throne! This TI helps to stabilize the generation results when you want someone sitting on the iron throne. Although the SD model is able to draw the iron throne in vanilla mode, the generation results might be drifted when you want some specific person sitting on it. This TI helps to mitigate the issue.

The inversion is trained based on Realistic Vision 2.0. It works under vanilla SD1.5, but is recommended to use under Realistic Vision 2.0.

Customized Training

Indulge in the ultimate fantasy and take your place upon the legendary Iron Throne. Immerse yourself in the captivating world of Game of Thrones as you grace this iconic seat of power. Each meticulously crafted blade of valyrian steel forms a mesmerizing backdrop, exuding an aura of authority and regality. Experience the thrill of embodying a ruler, capturing your essence atop this renowned symbol of dominion. For those craving an extraordinary adventure, send us an email, and let us transform your imagination into reality as you become a part of the storied legacy of the Iron Throne. If you are interested, feel free to contact us at genartai.site@gmail.com

Appendix

This inversion is trained with just 5 images with 1000 steps. More improvements are coming. Stay tuned!

Another of my attempts.

Again a thank you to Alyila for the guide.

again i doubt i have done a good job, but again i get reasonable results with azonya and clarity2.

Another of my attempts.

Again a thank you to Alyila for the guide.

again i doubt i have done a good job, but again i get reasonable results with azonya and clarity2. I may have done somehting off on this one as i do get some artifacts and marks in some results, so sorry if this one is a bust.

Another of my attempts.

Again a thank you to Alyila for the guide.

again i doubt i have done a good job, but again i get reasonable results with azonya and clarity2.

Another of my attempts.

Again a thank you to Alyila for the guide.

again i doubt i have done a good job, but again i get reasonable results with azonya and clarity2.

Another of my attempts.

Again a thank you to Alyila for the guide.

again i doubt i have done a good job, but again i get reasonable results with azonya and clarity2. Recommend you add to your prompt to help control the body if you want accuracy to the actual model.

One of my first attempts at a textual inversion, im not going to be good at this but wanted to be able to make some images with some of my favourite looks.

A big thanks https://civitai.com/user/Alyila/models for the guide that helped me, and also for making so many of my favourite SG's.

I dont think i have done a good job on this but i get reasonable results from Clarity2 and Azovya models.

Now this looks like a job for me
So everybody, just follow me
'Cause we need a little, controversy
'Cause it feels so empty, without me

taki from soul calibur

Brittni Rayne Whittington, popularly known by the name Whitney Wright, is a well-known American adult film actress and model.

Trained on 1.5 but examples are from Analog-madness

Skews nsfw, as you might suspect, but sfw is easy as well

Please let me know what you think. Drop me a line if you have comments/suggestions/recommendations/or requests for the next model.

23 year old American influencer and nude model.

A request of @argus2968.

Hailey Grice is an American social media personality with more than 1M followers on Instagram.

This is a 240-step TI trained on a dataset of 18 images with a full batch size (i.e., batch size 6 and gradient acc. 3). Once again, I used @Scottymac's free online Colab, so thanks again, pal. :)

Appreciate my work? My TIs are free, but you can always buy me a coffee. :)

Curious about my (usual) work process? I have summarized it here.

How to create a good prompt using my TIs

You're obviously free to experiment, but bear in mind that my TIs are trained with a more or less fixed phrasing, that normally starts with:

"photo of EMBEDDING_NAME, a woman"

So I recommend always starting your prompt like that and then building the rest of the prompt from there. For instance, "photo of beautiful (hgr1ce-step-240:0.99), a woman as a movie star, hair upsweep updo, sweater off-shoulders, trousers, at a movie premiere gala, dark moody ambience (masterpiece:1.2) (photorealistic:1.2) (bokeh) (best quality) (detailed skin:1.2) (intricate details) (nighttime) (8k) (HDR) (cinematic lighting) (sharp focus), (looking at the camera:1.1), (closeup portrait:1.1)"

This ti is a commission of @Zorglub

Kristin Kreuk is a Canadian actress. Kreuk gained fame for her portrayal of Lana Lang in the television series "Smallville," which aired from 2001 to 2011. After "Smallville," Kreuk went on to star in the television series "Beauty & the Beast," where she played the lead role of Catherine Chandler. In addition to her television work, Kristin Kreuk has appeared in various films such as "EuroTrip," "Street Fighter: The Legend of Chun-Li," and "Space Milkshake." She has also been involved in other television projects, including the miniseries "Ben Hur" and the series "Burden of Truth," where she plays the lead character and is also an executive producer.

Feel free to leave me a tip/buy me a coffee if you like what I am doing :)

This model is best used with weight 0.8-1.0. Your image might be blurred if the weight is too high.

Because this embedding is trained on photos, it tends to produce realistic faces. However, it also works well with anime style model, just use prompt like cute loli face to keep the 2d style.

note that the example images are generated with xformer, so the seed will not be deterministic.

Holly Peers (born 30 July 1986) is an English glamour model, born in Salford. She began modelling in 2009 when she was discovered by Alison Webster, photographer for the Sun newspaper, after appearing in the local press for a charity march through Manchester City centre to raise money for SOS (support our soldiers) Holly soon started appearing in The Sun's Page 3 (where she is known as Hollie Peers). She continues to model for Page 3, and she also regularly appears in magazines such as Nuts and Loaded.

She was also on the front cover of the 2011 and 2012 Page 3 calendars and the 2012 Hot Shots Calendar.

Peers was the highest new entry in "Nuts 100 Sexiest Babes 2010" at No. 6.

If you like my work and you feel like it, you can invite me to a Ko-fi!

My first try at a TI.

Do you want to make a Fiat Panda, and I don't mean the new one? AFAIK this is your only way at the moment. It's also my first time doing anything different from just generating images... so expect issues. They do kinda look correct, usually.

This embedding has been adjusted for SDHK02.safetensors

Free to use, call the model with this prompt: "ABYR2023II-8000"

Give me a like and credit to keep making more models. Download, share and have fun.

Buxom and shapely redhead stunner Christine Smith was born on April 6, 1979 in San Dimas, California. Christine was discovered by "Playboy" magazine when she accompanied a friend to a casting. Veteran photographer Stephen Wayda took some test photos of Smith, who was subsequently chosen by Hugh Hefner to be the Playmate of the Month in the December, 2005 issue of the famous men's magazine. She has posed for: Hooters calendar, Wilson Golf calendar, FireGirls, Muscle Magazine, Muscle Power and Nutrition, Easy Riders, Wicked Temptions, Atomix, Fever Magazine, and numerous others. Smith went on to appear in a couple of "Playboy" videos. She rehabilitates abused and unwanted animals. Christine has appeared in House Bunny, Bones, Last Comic Standing, CSI Miami, Celebrity Apprentice, Mad TV, Just Go With It, Walk Hard, Girls Next Door, Hancock, and recently Bad Teacher.

Trained on 1.5 base, but examples are on Analog Madness,

Skews NSFW as playmates tend to do, but SFW is easy too.

Defaults to redhead, but a few of the pictures were during her blonde phase. To ensure Redheadness, include redhead in your prompt.

If you have any comments/suggestions/criticism/suggestions for the next model please leave a comment or a review below.

My "first succesful" attempt that produce good faces.

It doesn't match the exactly the real person face I trained it on, but it produce some good faces quite consistent, when you use the same model.

Trained with a big dataset of 105 images at 13125 steps of the same real person.

We'll see if someone finds who => please guesses in private comments ;)

I didn't cheery pick a step number on purpose, you can try to use with lower steps which will produce slightly different results:

NoraHepsie-6700 [the rest of your prompt]

Hope you enjoy it !