
๐ Welcome to Stelra!
Stelra creates high-quality animated-style illustrations.
Create and enjoy AI-generated artwork with Stelra!
๐ Table of Contents
โ๏ธ Recommended Options
Clip skip : 2 โ
Sampling method : Euler a | DPM++ 2M Karras
Sampling steps : 20 ~ 40 โ
Width x Height : 512x512 โ
CFG : 6 ~ 10
Hires. fix : Enable
Upscaler : Latent
Hires steps : 20 โ
Denoising strength : 0.6 ~ 0.7
Upscale by : 1.5 โ
๐๏ธ Detail

Stelra
๐ฉ FAQ & TIP
Q. Can I publicly share the content created using that AI model?
A. Yes, it is possible, but you must disclose that the content was created using AI. Additionally, if you used Lora or similar techniques to apply the specific art style of a particular artist, it should never be publicly shared.
Q. I intend to use the model to generate malicious content targeting a specific individual.
A. It is absolutely not possible. Creating malicious content targeting specific individuals is strictly prohibited, both in private and public domains.
Q. I intend to use the model to create adult content. Are there any restrictions on this?
A. There are no restrictions on creating adult content using this model. However, all of the above must be observed, and the creation of adult content including children is strictly prohibited.
Q. The quality of the picture is appearing very low... Is it because the model's quality is too poor?
A. Please use a embedding model like 'easynegative' or enter the following content in the negative prompt.(worst quality:1.4, low quality:1.4, normal quality:1.4)
Q. The saturation of the outcome is low. Why is it like this?
A. Please use the VAE model. Apply various VAE models to find the one that suits your preferences!
Q. The pictures generated by the model appear slightly inappropriate... Please let me know how to resolve this issue!
A. This error occurs due to the use of an NSFW model during model merging. To prevent this error, please include the following content in the negative prompt.nsfw, hentai
Q. Is there a virus in that model? I'm very suspicious of the model!
A. The model itself does not contain any viruses. However, certain antivirus software may detect the model as a virus. To avoid such risks, use .safetensors instead of models with the .ckpt extension.
โน๏ธ Model Information
Stelra is a model created by appropriately combining existing Stable-Diffusion based models.
Below, you can find information about the models used in the creation of Stelra.
Models Merged into Stelra
- Counterfeit-V2.5
- Counterfeit-V3.0
- ๆฎๆฎ้้็่ๅWMW-1
- yden
- CoffeeMix
- pastel-mix
- anything-v3
- anything-v4
- 9527
- NostalYakiMix
- refgyoza
- AbyssOrangeMix3
- waifu-diffusion
- KawaiiMix
- NovelAI animefull-final-pruned
Model merging method
Tool : stable-diffusion-webui
Interpolation Method : Weighted sum & Add difference
Multiplier : 0.05, 0.25, 0.35, 0.4, 0.5, 0.65, 0.75, 0.95
I apologize, but there is no documented recipe for model merging, which means I cannot provide specific details about merging models, including AbyssOrangeMix3.
Some recipe
Primary Model : pastel-mix
Secondary Model : 9527
Merge Result : DrawX
Weighted sum | Multiplier : 0.35
Primary Model : AbyssOrangeMix3
Secondary Model : DrawX
Merge Result : DrawAI
Weighted sum | Multiplier : 0.4
Other merging operations...
๐ Licence
This model is open access and available to all, with a CreativeML OpenRAIL-M license further specifying rights and usage. Please read the full license here