Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,38 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
base_model:
|
4 |
+
- Laxhar/noobai-XL-Vpred-1.0
|
5 |
+
---
|
6 |
+
# シマムラアキト's PixelArt Style Model
|
7 |
+
|
8 |
+
This is the pixel art style model I used on X (Twitter), account: [@shimamuraa11451](https://twitter.com/shimamuraa11451). This model is designed to generate pixel-style images.
|
9 |
+
|
10 |
+
## 🔧 Recommended Settings
|
11 |
+
|
12 |
+
- **CFG (Classifier-Free Guidance Scale)**: 3
|
13 |
+
- **Step**: 20 \~ 30
|
14 |
+
- **Scheduler**: SGM\_UNIFORM
|
15 |
+
- **Sampler**: Euler series / DPM series
|
16 |
+
|
17 |
+
## 📏 Pixel Block Size & Rescaling Method
|
18 |
+
|
19 |
+
This model generates approximately `8×8` pixel blocks. To ensure proper alignment, you must use the **"NEAREST downsample to 1/8, then upsample back to the original size"** method.
|
20 |
+
|
21 |
+
- **ComfyUI**: This method can be implemented directly.
|
22 |
+
- **Automatic1111 WebUI**: Related plugins are available for automatic processing.
|
23 |
+
|
24 |
+
### Image Resolution Scaling
|
25 |
+
|
26 |
+
Since the model works with `8×8` pixel blocks, you should set the height (H) and width (W) according to the final pixel art resolution you want. Refer to the table below:
|
27 |
+
|
28 |
+
| Target Pixel Art Size | Input Resolution |
|
29 |
+
| --------------------- | ---------------- |
|
30 |
+
| 32×32 | 256×256 |
|
31 |
+
| 128×128 | 1024×1024 |
|
32 |
+
| 256×256 | 2048×2048 |
|
33 |
+
|
34 |
+
## ⚙️ Model Training Information
|
35 |
+
|
36 |
+
This model was fine-tuned using **noob vpred-1.0**.
|
37 |
+
|
38 |
+
Since it is vpred-based, additional steps may be required to ensure proper usage.
|