Turbocharge Your Portrait Game: 3x Faster Generation with Nunchaku!
1. Workflow Overview

This workflow optimizes commercial-grade portraits with:
3x Faster Generation: Uses
Nunchaku
quantized models.Skin Enhancement: Auto-fixes flaws via
W_skin.safetensors
LoRA.Tile Processing: Splits images to avoid VRAM overload.
Core Models:
svdq-int4-flux.1-dev
: Quantized model for speed.W_skin.safetensors
: Custom skin LoRA (weight=0.7).4x_NMKD-Siax_200k
: Upscales to 4K.
2. Key Nodes
Node | Function | Installation |
---|---|---|
| Loads quantized model | Install |
| Splits images into 1024x1024 tiles | Built-in |
| 4x upscaling ( | Manual model download required |
Dependencies:
Download models:
W_skin.safetensors
→ Save tomodels/loras/
.4x_NMKD-Siax_200k.pth
→ Save tomodels/upscale_models/
.
3. Workflow Groups
Input Image: Upload portrait (e.g.,
1d21adbf...png
).Model Load: Load Nunchaku model and skin LoRA.
Tile Sampling: Process high-res images in tiles.
Post-Processing: Upscale and reassemble tiles.
Output & Compare: Save 4K results with A/B comparison.
4. Inputs & Outputs
Inputs:
Image: ≥1024x1024 resolution.
Prompts: e.g., "fair skin, 8k photo".
Output:
4K portrait (saved to
ComfyUI/output
).
5. Notes
VRAM: ≥12GB GPU (8GB with tiling).
Debugging:
Missing models → Check
loras
andupscale_models
paths.
Optimization:
Set
FluxGuidance=3.5
for better skin details.Reduce
denoise=0.4
inBasicScheduler
for cleaner results.