Update README.md
Browse files
README.md
CHANGED
|
@@ -24,6 +24,8 @@ This is a lightweight model designed to remove JPEG compression artifacts, digit
|
|
| 24 |
|
| 25 |
> **Which version to pick?** V2 generally restores better, but it is also cautious and may decide an image is already clean. If V2 feels too conservative for a specific image, try the V1 weights (`best_ema_15E.safetensors`) and keep the result you like more.
|
| 26 |
|
|
|
|
|
|
|
| 27 |
## Examples
|
| 28 |
|
| 29 |
|
|
|
|
| 24 |
|
| 25 |
> **Which version to pick?** V2 generally restores better, but it is also cautious and may decide an image is already clean. If V2 feels too conservative for a specific image, try the V1 weights (`best_ema_15E.safetensors`) and keep the result you like more.
|
| 26 |
|
| 27 |
+
> **Force processing (ComfyUI node):** If V2 decides an image is already clean, enable `force_process` and raise `force_noise_std` (default 0.02; 0.05 for stubborn cases). This injects tiny noise so the model always runs, while the final output is still blended with the original.
|
| 28 |
+
|
| 29 |
## Examples
|
| 30 |
|
| 31 |
|