Kristanna Sommer Loken is a Norwegian-American actress best known for portraying the T-X Terminator in Terminator 3. She has also acted for a few other projects but by the looks of it, it seems best they are left entirely unmentioned. I decided to embed her after making a LORA of Kristanna as the T-X.

The embed has been trained on SD 1.5 and works well on most 1.5 based models. I've noticed with some models it might be necessary to use increased token weight to get backgrounds or settings to render properly.

If you enjoy my creations or have a face you'd like to place a commission, consider supporting me on Ko-Fi :)

Embedding of actress Maggie Siff, known from series such as Mad Men, Sons of Anarchy and Billions. One of my favorite actresses with a strong character and expression. I'm happy to say this one turned out great. Her likeness is easily achieved with this TI.

DPM++ 2M KARRAS 30 steps

7 CFG

Standard weight is fine.

Upscale through hires fix or with Contronet 1.1.

Renèe Estella is a model and social media personality with more than 800K followers on Instagram.

1000-step TI trained on a dataset of 18 images with my usual settings.

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

Curious about my 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 (r3nestella:0.99), a woman, RAW, close portrait photo, bodysuit with zipper, cleavage, fishnet stockings, pale skin, slim body, (high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3 sharp focus, f 5.6".

I trained this model with 10.000 steps (v1.0) and 20.000 steps (v1.1), I'm no expert, but I hope, you like it.
Add into your prompt: "(ElviraMistress:0.75)"
Trained with 512x512 pixels.

I accidentially deleted version 1.0, re-uploaded, sorry :)

Yet Another horror style on Civitai :)

About

A style for generating dark (visually an thematically) Gothic Horror style pictures. This was put together using the Embedding Inspector addon. I combined visually and thematically dark vectors into an embedding with weightings I liked. This style doesn't do anything special, other than guide your checkpoint to a horror themed result.

If you're an experienced prompt crafter, you probably don't need this, and could generate similar results with out it.

Installation

Place in the directory where you keep your embeddings

Usage

Use style-gothic-horror in your prompt. Placement location in the prompt really matters with this one. Placement at the beginning of the prompt creates visually dark images, (this works with noise offset lora, or models that include noise offset). Placement at the end of the prompt will give you more visual flexibility. Experiment with placing it at the beginning, middle or end of your prompt.

This is a Textual Inversion i've made for fast photo-realism.

It comes with positive and negative TI. Use it with weight do difference.

Download all 2 files from "files option" in top right.

Put in embeddings folder and use it like any other embed.

Usage:

Prompt: your prompt, kkw-ph1

Negative prompt: kkw-ph1-neg

Don't know why civitai striped out from data.

You also can use the negative as standard for all other type of work.

Use it in img2img to port 3d render and other work to real

just put

Prompt: subject, kkw-ph1

Negative prompt: kkw-ph1-neg

Set cfg to 30 and denoise lower( 0.18) for test, then move it for get better result.

This setting is for image like Daz3d image, if you use anime or really flat image, you have to go up with denoise value.

Also remember more high is denoise, more lower you need cfg to keep image consistance.

测试完成了。接下来我会弄2.0版本的，全新的训练集处理和新的参数

The test is complete. Next, I will work on version 2.0, new training set processing, and new parameters

更正一下，trigger name 是我乱填的 a woman

真正的tirgger name 是embedding本身的名字 我已经更正了。为这个失误抱歉

Correct it, the trigger name is a woman that I filled in incorrectly

The real tiger name is the name of embedding itself. I have corrected it. Sorry for this mistake

Don't underestimate emb by it's only over a dozen kb. Although this metaphor is not appropriate, early computers were as big as houses, but how big are chips today?

In a way embedding is better than lora with face loyalty restoration, hope more bro try to train your emb.

Onika Tanya Maraj-Petty (born December 8, 1982), known professionally as Nicki Minaj, is a Trinidadian-born rapper, singer, and songwriter based in the United States. She is known for her musical versatility, animated flow in her rapping, alter egos and accents. Minaj first gained recognition after releasing three mixtapes between 2007 and 2009. Her debut album, Pink Friday (2010), topped the U.S. Billboard 200 chart. Its fifth single, "Super Bass", reached number three on the U.S. Billboard Hot 100 chart and was certified diamond by the RIAA. Minaj's follow-up album, Pink Friday: Roman Reloaded (2012) explored dance-pop. The lead single, "Starships", peaked in the top five worldwide.

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

Em gái Việt ngây thơ - Innocent beautiful Vietnamese girl.

Trained using v1-5-pruned-emaonly

This person does not exist

Sample images are using Dynamic Thresholding set to 4, see the third image for a comparison for the input CFG effects.

Lindsay Czarniak

Lindsay is the most beautiful sports anchor that has ever existed.

You can learn more about my process in my blog post HERE

I can create a custom Textual Inversion or LoRA just for you!

Do you like my work? Buy me a beer!

[ SDv1.4, 4e-03, 4600 steps, v/s .200 ]

Carla Poquin alias Poqssi, French actress

Trained using v1-5-pruned-emaonly

This person does not exist

Sample images are using Dynamic Thresholding set to 7, see the third image for a comparison for the input CFG effects.

Buse Yıldırım is a Turkish news anchor and can also be found on Instagram.

She has over 80k followers.

Still learning and improving the progress of the training, any feedback is welcome.

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

An embedding of actress Hilarie Burton, mostly trained on images of her One Three Hill days. It can also serve as an embedding for the character of Peyton Sawyer, otherwise known as many teen boys' wet dream in the early 2000's.

512x512 or 512x768

For some reason, it's not easy to get good likeness at low weight. If you want to optimalize likeness I would put the weight of the embedding at 1.05 to 1.1, depending on the results you get.

30 steps DPM++ 2M KARRAS

CFG 6-7

Charlie can be found Instagram. Done as a request for a user who wishes to be anonymous. I hope this ti is what you were looking for!

Still learning and improving the progress of the training, any feedback is welcome.

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

Michelle Lim Davidson is an Australian actress of South Korean decent. You can find her Instagram page here.

Still learning and improving the progress of the training, any feedback is welcome.

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

A commission of @Zorglub.

Shania Twain is a Canadian singer and actress known as the Queen of Country Pop. She's one of the best selling musicians of all time and a winner of 5 Grammy Awards.

(Update: Kept training the model and testing some steps, and 1800 seems better than the previous one I uploaded, so here it is.)

1800-step TI trained on a dataset of 18 images with my usual settings.

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

Curious about my 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 (shtw41n:0.99), a woman, RAW, close portrait photo, leopard print suit with hood, leopard print top, pale skin, slim body, (high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3 sharp focus, f 5.6, redhead, (hood on:1.2)".

26 year old Ukrainian cosplayer.

_________________________

Feel free to support me.

Ece Belen is a turkish news anchor and can also be found on Instagram.

She has over 100k followers.

Still learning and improving the progress of the training, any feedback is welcome.

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