a semi-Realistic(2.5D) NoobAI SDXL v-pred based Mixed model

This is another open/free merged model named Evermore SDXL v-prediction model.

NoobAI XL (NAIXL) vpred model is developed by Laxhar lab @huggingface

A v-prediction model requires some special config settings (distinct from normal eps-prediction model, you may need to use comfyUI or reforge or even A1111 dev branch. (please see also https://huggingface.co/Laxhar/noobai-XL-Vpred-1.0 for details)

for A1111 user

• you need to swtich dev branch

• v_pred and ztsnr (Zero Terminal SNR) will be automatically switched on

• for some v_pred models without v_pred and ztsnrstate_dict keys, you may need to add these keys manually by python script (please see https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16567) or you need to copy sd_xl_v.yaml yaml config into the models/Stable-Diffusion dir with the same name of satetensors (for example, foobar.safetensors needs foobar.yaml and check Zero Terminal SNR setting in the A1111 UI settings.

Version summary

• v0.7 - experiment v-pred model

Merged Models

• NoobAI SDXL v-pred model by @L_A_X

• NoobReal by @Sennke - add block level merge (a traininig model)

• evermore illustrious XL by @mixboy - block level merge for backgrounds, realistic skin, etc.

Recipes

recipe details are included in the model checkpoint or some images. (so you can use/modify this recpie easily by using the Model-Mixer A1111 extension)

• Fp16 fixed VAE included.

• all recipes of each version will be included in the checkpoint metadata!

Recommends

• All posted images use μ-DDetailer script extension

• recommended sampler: Euler a + scheduler: SGM Uniform etc.

  • note: NoobAI XL vpred recommends Euler

• recommended image size (width range is about 640~1024 / height range is 640~1344):

  • 1024x1024 (default SDXL)

  • 640x960 / 768x1152 / 800x1200

  • 832x1216 / 896x1152

  • see also SDXL resolution cheat sheet: https://www.reddit.com/r/StableDiffusion/comments/15c3rf6/sdxl_resolution_cheat_sheet/

• recommended CFG scale: 3.5~6.5

• Set Clip skip: 2 recommended (Clip skip:1 also works fine), ENSD: 31337

Useful AUTOMATIC1111's webui extensions

the following sd-webui extensions are recommended.

• Model-Mixer extension

• μ-DDetailer script extension

  • or use famous ADetailer extension

• Dynamic Thresholding (CFG Fix) extension

• SuperMerger extension

• Lora Block Weight extension

• civitai extension to add several useful features for civitai.

ChangeLog

• 2025/02/01 - first release v0.5

TODO

• add more realistic skin tone

• optimize details adjust parameters for more realistic with high details

• more asian style face

• fix detail levels

License

Fair-AI-public-1.0-sd - This license disclaim "Same license restriction" so I do not add any restriction except same license restriction.

This model permits users to:

✔Use the model without crediting the creator

✔Sell images they generate

✔Run on services that generate images for money

✔Share merges using this model

✔Sell this model or merges using this model

❌Have different permissions when sharing merges

Fair-AI-public-1.0-sd license is a “SD variant” license contains a Prohibited Uses section designed to be compatible with Stable Diffusion’s license.

All Ilustrious XL based training models are distributable under the same license, Fair-AI-public-1.0-sd, and this license compatible with the original SD license.

as described in the Fair-AI-public-1.0-sd lincense, you can also use any non-model parts of this software under the terms of the GNU AGPL 3.0, or any later version of that license.
If you do, No Harm and No Liability still apply.

the OpenRAIL-M license permits users to own the rights to the images they generate and to use these images for commercial purposes​ (Stable Diffusion)​​ (Baseten)​. This openness supports a wide range of applications, from creative projects to commercial services, enabling businesses and individuals to leverage the model's capabilities for various purposes.

Therefore, if you're considering using Stable Diffusion for commercial products or services, the licensing terms do support such use, as long as the guidelines and restrictions outlined in the license are followed. (from chatgpt)

Support me

If you like my work, feel free to buy me a coffee at ko-fi. https://ko-fi.com/mixboy