🗡️💀 DaSiWa-Illustrious-XL | Realistic 💀🗡️
Version overview: https://civitai.com/articles/23495/dasiwa-model-versions-and-timeline
This time a tried a couple of new techniques to implement more features and coherence without burning the weights to produce a Illustrious realistic model. On my testing this works very well overall, but some details need more or less refinement with a detailer.
The overall comparison to other realistic models was very good (imho) and sometimes surpasses them significantly on limited testing.
🔮 Key Features:
🔥 LoRA-Free Generations
Generate high-quality art without stacking detailers (unless you want adding spacial styles/concepts).Support of IL 2.0 features
Consistent characters
Extreme versatile (use quality prompts to get what you want)
Quality motions
🔞 NSFW:
Enhanced anatomy + poses + framing
NSFW and SFW capable
For the love to 🐈⬛-Girls
Stable resolution support up to 1536x1536p without upscaling.
🎭 Crisp Details: Enhanced details, poses, backgrounds, outfits, and character features (all of them 🖤)
🪄 Prompt Responsiveness (play with CFG (5-7).
📜 VAE Baked in
Read version details for the version you are using for more information!
🛠️ Recommended Settings
Steps: 25-30 (I use 30 normally)
CFG: 5-7 (I use 5 or 6 normally)
Sampler: Euler
Scheduler: Normal or Simple
Resolution up to 1024x1536 (native quality).
Standard Illustrious & SDXL resolutions also supported (1024 x 1024, 1152 x 896, 896 x 1152, 1216 x 832, 832 x 1216, 1344 x 768, 768 x 1344, 1536 x 640, 640 x 1536).
My go to settings:
Aspect ratio 3:4 (basic images) or 9:16 (for i2v images)
CFG 5
Euler/Normal
30 steps
🩺 Fixes & Feedback
Some details (like eyes, finger) may need a detailer: Use a detailer, or inpaint with mask radius 30-60 and 0.2-0.3 strength over it.
If you use LoRAs, try to respect the LoRA training resolution, some LoRAs are not trained for high output and will show artefacts - than use inpaint to correct that.
🖤 Why I Made This
I was tired of using all these embeddings, detailers, negative and quality prompting just to get a good image. Quality prompting is only used to get a specific result, not because the image will get trash without it.
So I can just make my images with some style/character LoRAs without pushing 5-10 LoRAs into a generation.
This checkpoint is also my personal playground.
Closing words
🐞Recommended offline Frontend/UI: SwarmUI
I want to thank all the fantastic other creators who made super nice merges to play with!
YOU are responsible for outputs as always! If you make ToS violating content and I get aware I WILL report this.
Disclaimer
This models are shared without warranties and with the condition that it is used in a lawful and responsible way. I do not support or take responsibility for illegal,
Custom License Addendum: Distribution Restriction
Notice: Notwithstanding the base license selected for this model, the following restrictive terms apply:
No Redistribution: You are not permitted to host, mirror, or redistribute this model (checkpoint, LoRA, or Safetensors files) on any other platform, website, or service (including but not limited to Hugging Face, Tensor.art, or SeaArt) without explicit written permission from the creator.
Attribution & Source: This model is officially maintained only on Civitai or other platforms where I explicitly own the repository. To ensure users receive the correct version, updates, and safety metadata, please point users to the original URL.
Usage: All other rights regarding the use of the model for image generation remain as per the terms and the restrictions provided per model.