Models>

Fares Fares - Textual Inversion - SD 1.5

Quick Generate
Select a style, write what you want to see and click Generate
What you want in the image.
Clear
Random
* This model is not available for use and might generate different characters compared to the ones seen in the images.

Model Description by Creator

Use weight 0.92 / don't use on diffuser backend on SD.Next

Fares Fares (born 29 April 1973) is a great Swedish-Lebanese actor of Assyrian descent. Known for his performances in the video game A Way Out (2018), TV series Westworld (2018), Chernobyl (2019) and The Wheel of Time (2021).

.

.

I really don't like TI. I don't intend to train it anymore. It's hard to get to resemblance and it looks OKish at best compared to Lora.

It can be worth to use together with another lora to "stabilize" the other loras changes and rigidness.

For example: if I want Fares Fares to have a pink colored hair and a really long beard, the Fares Fares SD15 lora is hard at that because it knows Fares Fares don't have those characteristics (even using a Leco slider Lora). So using the embedding apparently breaks that a little.

But as a trade off it will make the final result more "standard". For example with fantasy loras, like the Venomancer, the SD Lora alone it's way more flexible and play together nicely. It's more "fantasy like" than this TI alone or they combined.

Also embedding + lora can SOMETIMES increase resemblance. Some other times it will get crappy and saturated. It's inconsistent but it happens, I'll post some examples.

So for my recommendation is:

TI alone: 0.92

TI + Fares Fares Lora: TI:0.92 + Lora:0.8

0.92 Yes it's very specific. You can lower it if its too much saturated, but will lose resemblance. Use saturated color on the negative prompt.

For this SD15 TI I did 20 (!) training versions, they all looked like crap, until I realized that SD.Next diffuser backend simply don't work well with safetensors embedding, I guess. SO don't use diffuser backend for SD15 things, it's not good. The diffuser backend is super fast but even with normal Loras, it gets weird.

I feel like using the prompt keyword "Fares Fares" is not needed. But for using together with Fares Fares lora, might be a good idea to use it at "(Fares Fares:0.8)"

There is no Keywords, use the name of the embedding:

(Fares_Fares_TI_SD15:0.92)

  

Negative supporting prompts:

glasses, saturated colors, dim lit, shadow face

 


Try out my other SD 1.5, SDXL and TI here:

Fares Fares – Lora SD15

Fares Fares – Lora SDXL

Fares Fares – TI SDXL (will update later)


The base checkpoint is the realisticVisionV51_v51VAE 3.97GB. I used it instead of the original 1.5 because this is not a broad concept. To get best resemblance you should use the best model for it, so use realistic vision 5.1.

Please use high-res fix to get better results

CFG: 5.5

Clip skip: 1

512px

 

DPM++ 3M Exponential (50 steps or more)

DPM++ 2M Karras (25 Steps or more)

DPM++ 2S a Karras

 

Current problems with the Lora:

  • Saturate colors at high weight

  • loose resemblance if below 0.8

 

Some settings: Trained 768res. 177 pictures. Token name "Fares Fares". Initialization word "man".

Because I did so many versions (20) and the best one was the second and Kohya setting on TI "Training comment" is not saved I don't know how many steps this was. I THINK it was 10 steps for best pictures, 5 steps for OK pictures. 2 different folders (sharing the same “Fares Fares person" concept).

I used AdamW, constant, 12 epochs. Gradient accumulate steps at 1. Batch size 1.

It would be around 16.620 steps I think.

Hopefully you can leave some results and some comments. Any idea is appreciated. Thank you.

Images Generated With This Model

(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, wearing viking clothes, bloody face, front view, in the woods forrest, middle-aged daddy, skin detail, pores, (messy:1.2) medium long hair, studio light, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
N/A
Hidden For Public Users
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, wearing viking clothes, bloody face, front view, in the woods forrest, middle-aged daddy, skin detail, pores, (messy:1.2) medium long hair, studio light, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, portrait photo front view, on a movie premiere, happy, messy medium length hair, 50 y.o, scarf, glasses, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, closed mouth, lips, (beautiful eyes, iris, pupil:1.1), face birthmark
N/A
Hidden For Public Users
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, portrait photo front view, on a movie premiere, happy, messy medium length hair, 50 y.o, scarf, glasses, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, closed mouth, lips, (beautiful eyes, iris, pupil:1.1), face birthmark
photo portrait of Fares Fares (Fares_Fares_SD15_v2-012:0.92) man a 50 years old man floating in his clothes (underwater:1.3) surrounded by colorfull fishes, happy, natural (sunny:1.3) lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
N/A
Hidden For Public Users
photo portrait of Fares Fares (Fares_Fares_SD15_v2-012:0.92) man a 50 years old man floating in his clothes (underwater:1.3) surrounded by colorfull fishes, happy, natural (sunny:1.3) lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
(Fares Fares:1.2) (Fares_Fares_SD15_v2-012:0.92) man with a (slicked-back hair:1.4) wearing a black vest (and long sleeved white shirt:1.25), inside a dark cave, (full:1.2) (long:0.5) (squarish:1.2) big (beard:1.2), devilish (confident looking down:1.2), (a ven0mancer, venom liquid, swirling green liquid:1.1) is around his body, 4k uhd, dslr, soft light, high quality, Fujifilm XT3 <lora:weight_slider_v2:-0.5> <lora:detail_slider_v4:0.92> <lora:Concept - Venomancer:0.35> <lora:muscle_slider_v1:-0.8>
N/A
Hidden For Public Users
(Fares Fares:1.2) (Fares_Fares_SD15_v2-012:0.92) man with a (slicked-back hair:1.4) wearing a black vest (and long sleeved white shirt:1.25), inside a dark cave, (full:1.2) (long:0.5) (squarish:1.2) big (beard:1.2), devilish (confident looking down:1.2), (a ven0mancer, venom liquid, swirling green liquid:1.1) is around his body, 4k uhd, dslr, soft light, high quality, Fujifilm XT3 <lora:weight_slider_v2:-0.5> <lora:detail_slider_v4:0.92> <lora:Concept - Venomancer:0.35> <lora:muscle_slider_v1:-0.8>
(Fares Fares:1.3) (Fares_Fares_SD15_v2-012:0.92) portrait photo of a man, front view, on a movie premiere, happy, messy medium length hair, age:50, scarf, glasses, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, closed mouth, lips, (beautiful eyes, iris, pupil:1.1), face birthmark
N/A
Hidden For Public Users
(Fares Fares:1.3) (Fares_Fares_SD15_v2-012:0.92) portrait photo of a man, front view, on a movie premiere, happy, messy medium length hair, age:50, scarf, glasses, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, closed mouth, lips, (beautiful eyes, iris, pupil:1.1), face birthmark
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, front view medium body shot, age 50, wearing a soccer uniform, stylish quiff hair, suit, mustache only beard, laughing, natural lighting, sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, (tongue:0.5)
N/A
Hidden For Public Users
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, front view medium body shot, age 50, wearing a soccer uniform, stylish quiff hair, suit, mustache only beard, laughing, natural lighting, sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, (tongue:0.5)
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, portrait photo front view, on a windy beach, nice looking, messy medium-length bluehair, 50 y.o, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, (closed mouth:1.2), lips, (beautiful eyes, iris, pupil:1.1)
N/A
Hidden For Public Users
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, portrait photo front view, on a windy beach, nice looking, messy medium-length bluehair, 50 y.o, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, (closed mouth:1.2), lips, (beautiful eyes, iris, pupil:1.1)
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, portrait photo front view, on a windy beach, nice looking, messy medium-length bluehair, 50 y.o, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, (closed mouth:1.2), lips, (beautiful eyes, iris, pupil:1.1)
N/A
Hidden For Public Users
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, portrait photo front view, on a windy beach, nice looking, messy medium-length bluehair, 50 y.o, skin detail, natural lighting sunlight, 4k uhd, dslr, soft light, high quality, Fujifilm XT3, long beard, (closed mouth:1.2), lips, (beautiful eyes, iris, pupil:1.1)
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, wearing viking clothes, bloody face, front view, in the woods forrest, middle-aged daddy, skin detail, pores, (messy:1.2) medium long hair, studio light, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
N/A
Hidden For Public Users
(Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, wearing viking clothes, bloody face, front view, in the woods forrest, middle-aged daddy, skin detail, pores, (messy:1.2) medium long hair, studio light, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
(viking:1.3) (Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, (wearing viking clothes:1.4), front view, in the woods forrest, middle-aged daddy, skin detail, pores, (messy:1.2) medium-length long hair, studio light, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3
N/A
Hidden For Public Users
(viking:1.3) (Fares Fares:0.8) (Fares_Fares_SD15_v2-012:0.92) man, (wearing viking clothes:1.4), front view, in the woods forrest, middle-aged daddy, skin detail, pores, (messy:1.2) medium-length long hair, studio light, natural lighting, 4k uhd, dslr, soft lighting, high quality, Fujifilm XT3

Train Models

Train AI Models On Your Own Pictures

You can use your own pictures to train AI models. Reimagine yourself in new ways with our AI-powered image generator.
Create pictures of yourself around the world, in your dream outfit, as a video game character, or in countless art styles. The possibilities are endless.
Train Your Own Model

Generate AI Images. Simple and Easy

AIEasyPic is the #1 AI image generator. Generate images with simple prompts, swap faces, train models to reimagine yourself, and more.