Instructions to use Bl4ckSpaces/Ilustrious-Nova-V2-BlockMerge with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Bl4ckSpaces/Ilustrious-Nova-V2-BlockMerge with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Bl4ckSpaces/Ilustrious-Nova-V2-BlockMerge", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
Ilustrious-Nova-V2-BlockMerge
V2 Advanced Merge. This model uses SLERP (Spherical Linear Interpolation) with Block-Weighted merging strategy to combine the best of IllumiYume and NoobAI.
The "Pro" Recipe
We use a gradient merging strategy to preserve NoobAI's superior anatomy while keeping IllumiYume's beautiful rendering.
- Method: SLERP (Prevents contrast loss/washed out look)
- Base: IllumiYume XL v3.5
- Target: NoobAI XL v1.0
Layer Configuration:
- Input Blocks (Structure): ~40% NoobAI (Better posing/anatomy)
- Middle Block: 50% Mix (Balanced core)
- Output Blocks (Style): 10-20% NoobAI (Preserves IllumiYume's vibrant lighting & coloring)
Recommended Settings
- Scheduler: Euler a / DPM++ 2M SDE Karras
- Steps: 28-35
- CFG: 5-7
- V-Prediction: Enabled
- Downloads last month
- 2