Readme
Inputs
| Name | Type | Description | Default |
|---|---|---|---|
media |
file / URL | Video (.mp4, .mov) or image (.png, .jpg, .webp) to restore. |
– |
cfg_scale |
float | Classifier-free guidance scale (higher = stronger restoration). | 1.0 |
sample_steps |
integer | Diffusion sampling steps (1 = fast one-step mode). | 1 |
sp_size |
integer | Sequence-parallel shard heuristic (single-GPU build only accepts 1). | 1 |
fps |
integer | Frames-per-second for video outputs. | 24 |
seed |
integer | Random seed. Use -1 for random seed each call. | -1 (random) |
output_format |
string | Image output format (only used for image inputs). | png |
output_quality |
integer | Image quality for lossy formats (jpg/webp). | 90 |
apply_color_fix |
boolean | Apply optional wavelet color correction (matches official demo). | false |
model_variant |
string | Model size to run. | 7b |
resolution |
integer | Target resolution for the shortest edge in pixels. Automatically maintains aspect ratio. | 1080 |
max_resolution |
integer | Maximum resolution limit for any dimension (0 = no limit). | 3840 |
Model created