A.Ikari - SFW&NSFW

CHECKPOINT
Original


Updated:

12K

Introduction to the model:

This model is a result of a challenge or commitment i took to create a 2D Sd1.5 anime model which can generate better hands. Of course that is in comparison to my previous models.

This model took a while but is promising or you can say an upgraded version of my older IlluXionMix_v1 Illustral. I learned a lot in the process, explored many things, tried some of them and got decent results.

I won't promise or guarantee about its hand generations would be accurate all the time, but yeah it works pretty good for me. It can create good hands in complex designs also as you can see in the preview tab, Read below for the generation parameters and feel free to experiment things on your own.

<------------------------------------------------------------------------------------------->

In some cases it may need a facedetailer if your character is too far in the background otherwise the closeup results are absolutely great.

NOTE:- any Vae is not baked in particular use it on your own preference. Or you can use this without vae also, i mean the colours are pretty average if you are curious to know, it will not become white or faded like other models.

Hires.fix is Highly recommended to use to get better result.

Recommended Parameters:

*Sampler:- DPM++ SDE Karras, DPM++ 2M karras, Euler (I have tested these and got good results, feel free to play with your own sampler)

*Steps:- 20 is minimum and after 20 its on your choice (Average Steps usage is between : 20 - 50)

*CFG Scale:- 4 to 12 ( 6 or 7 is generally used )

*Vae:- Blessed2.vae, animevae, clearvae (Blessed2 is the one i generally use because it has a rather dark complexion)

*Upscaler:- R-ESRGAN 4x+Anime6b, 4x-ultrasharp

*Hires Steps:- Use half of the steps you have used to generate the image in the hires steps

(Example : If you are using 30 steps for generating the image use 12 to 15 steps in the hires.fix)

*Denoising:- 0.3 to 0.7 of Denoising

*Clip Skip:- 2

*Embeddings:- easynegative, FastNegativeV2, badhandv4, ng_deepnegative_v1_75t, verybadimagenegative_v1

Socials:

Follow me on Instagram:- https://www.instagram.com/itz.blitzz_/

Follow me on Pixiv:- https://www.pixiv.net/en/users/99137352

Follow my patreon page:- https://www.patreon.com/user?u=108322161

Problems You may face:-

Troubleshoots:-

Why are the colours in my image faded?

You have not used a VAE. For Good balanced colours i would recommend Blessed2 for this model (It worked the best for me). Feel free to explore and experiment different VAE

Why are my style lora's not working or giving me the desired result?

You may have used a negative embeddings which is altering the result.

Try removing that negative and generate it again

Support:-

If you face any problem regarding this model you can contact me on Instagram.

You can also download my model on:

Civitai: https://civitai.com/models/225392/aikari-sfwandnsfw

Version Detail

SD 1.5
Vae Not Baked Recommended: Blessed2

Project Permissions

Reprinting is strictly prohibited

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Comments

Related Posts

Describe the image you want to generate, then press Enter to send.