unico / README.md
Cinnamomo's picture
readme update; x/y plot will be updated asap
bf87e71
|
raw
history blame
No virus
4.34 kB
---
language:
- en
tags:
- stable-diffusion
- text-to-image
license: creativeml-openrail-m
inference: true
pipeline_tag: text-to-image
---
# i modelli 'Unico'
<img src=https://i.imgur.com/5KfDOik.png
width=100% height=100%>
Unico is the series of custom mixed models. Based on Inizio Unico and AbyssOrange2 models with U-Net Merge, and support .safetensors format only.
[WebUI](https://github.com/AUTOMATIC1111/stable-diffusion-webui)-amicable
## Summary
This model repository includes 4 main models currently:
1.
| Model: A | Model: B | Merge Weight | Base alpha | Merge Name |
| --- | --- | --- | --- | --- |
| [Inizio Fantasma+Inizio Inseguitore+Inizio Foschia](https://huggingface.co/Cinnamomo/inizio) | [Inizio Replicante+Inizio Skinjob+Inizio Deckard](https://huggingface.co/Cinnamomo/inizio) | weighted, M=0.66666666+M=0.66666666 | N/A | *Unico* |
Unico is another form of [Inizio Unico](https://huggingface.co/Cinnamomo/inizio).
2.
| Model: A | Model: B | Merge Weight | Base alpha | Merge Name |
| --- | --- | --- | --- | --- |
| [Inizio Unico](https://huggingface.co/Cinnamomo/inizio) | [AbyssOrange2 SFW](https://huggingface.co/WarriorMama777/OrangeMixs) | weighted, M=0.75. | N/A | *Unico Arancia* |
Unico Arancia('Orange🍊') is the closest model from AbyssOrange2 SFW.
3.
| Model: A | Model: B | U-Net Merge Weight | Base alpha | Merge Name |
| --- | --- | --- | --- | --- |
| Unico Arancia | ([Openniji](https://huggingface.co/Korakoe/OpenNiji) | 0,0,0,0,0,0,1,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 | 0 | *Unico Arancia 1.3* |
Unico Arancia('Orange🍊') 1.3 is improved model of Unico Arancia. Anime~Semi-realistic
4.
| Model: A | Model: B | U-Net Merge Weight | Base alpha | Merge Name |
| --- | --- | --- | --- | --- |
| Unico Arancia | ([Openniji](https://huggingface.co/Korakoe/OpenNiji) | 0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0 | 0 | *Unico Arancia Speciale* |
Unico Arancia Speciale('Orange🍊 Special') is specially improved model of Unico Arancia to increase delicate details.
5.
| Model: A | Model: B | U-Net Merge Weight | Base alpha | Merge Name |
| --- | --- | --- | --- | --- |
| (Unico Arancia+Unico Pera+Unico Cartamo+Unico Castagna+Unico Vaniglia) | (Elldreth 4060+SamDoesArtsUltMerge) | experimental mix | N/A | *Unico Castagna 2.0* |
Unico Castagna('Chestnut🌰') 2.0 is the improved model which focused to Semi-realistic only.
- NOTE: Another models are moved to legacy folder.
## Setting Reccomendation
```
##Basic prompts for anime
"txt2img/Prompt/value": "(best quality, extreme intricate detailed, octane render, very delicate cinematic light, colourful), (/*place tags*/), (solo girl/*character tags*/), (/*pose tags*/), (big breasts, big pelvis, slim waist, long legs, best ratio four finger and one thumb, /*body tags*/), (beautiful eyes and face), (/*colour of hair tag*/ hair, /*colour of eyes*/ eyes, thick lips, lip gloss), (/*clothing tags*/)",
"txt2img/Negative prompt/value": "(nsfw, worst quality, low quality:1.4), (greyscale), (fingers(missing, fused, interlocked, abnormal, too many, bad anatomy, fused, fusion, lose, bad detailed, mutated), digit(extra, fewer), hands(greater than 4 fingers, less than 4 fingers, cropped, mutated):1.4), (fat, chubby, curvy, watermark, signature:1.4), (3d, realistic)"
```
- Variational Automatic Encoder: [SD MSE 840k.vae.safetensors](https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.safetensors)
- Clip Skip: 2
- Resolution: 1024x576 w/ HighRes. Fix
- HighRes. Fix: R-ESRGAN General WDN 4xV3; upscale by 1.25
## Sample Images
>
> <img src=https://i.imgur.com/4VUj0fs.jpg
> width=100% height=100%>
> ▲ X/Y Plot #1
>
> <img src=https://i.imgur.com/aYIyVFJ.png
> width=100% height=100%>
>
> <img src=https://i.imgur.com/pKNd2XO.png
> width=100% height=100%>
>
> <img src=https://i.imgur.com/GknH4e0.png
> width=100% height=100%>
>
> <img src=https://i.imgur.com/rVblL4d.png
> width=100% height=100%>
> ▲ Unico Arancia
## License Information
This model follows Creative ML Open RAIL-M: [Stable Diffusion License](https://huggingface.co/spaces/CompVis/stable-diffusion-license)
But, You may use whatever you want. I don't like to set such restriction.
## Contact
*[email protected]* or [*Find Cinnamomo on Arcalive AI Art Channel*](https://arca.live/b/aiart)