You're looking at a specific version of this model. Jump to the model overview.
zsxkib /prototype-model:d52ae725
Input schema
The fields you can use to run this model with an API. If you don’t give a value for a field its default value will be used.
Field | Type | Default value | Description |
---|---|---|---|
input_path |
string
|
videos/pexels-cottonbro-studio-6649832-960x506-25fps.mp4
|
Input Video
|
prompt |
string
|
white ancient Greek sculpture, Venus de Milo, light pink and blue background
|
Prompt
|
image_resolution |
integer
|
512
Min: 256 Max: 512 |
Frame resolution
|
control_strength |
number
|
1
Max: 2 |
ControNet strength
|
color_preserve |
boolean
|
True
|
Preserve color (Keep the color of the input video)
|
left_crop |
integer
|
0
Max: 512 |
Left crop length
|
right_crop |
integer
|
0
Max: 512 |
Right crop length
|
top_crop |
integer
|
0
Max: 512 |
Top crop length
|
bottom_crop |
integer
|
0
Max: 512 |
Bottom crop length
|
control_type |
string
(enum)
|
HED
Options: HED, canny |
Control type
|
low_threshold |
integer
|
50
Min: 1 Max: 255 |
Canny low threshold (If `Control type` is "canny" Control type)
|
high_threshold |
integer
|
100
Min: 1 Max: 255 |
Canny high threshold (If `Control type` is "canny" Control type)
|
ddim_steps |
integer
|
20
Min: 1 Max: 20 |
Steps (To avoid overload, maximum 20)
|
scale |
number
|
7.5
Min: 0.1 Max: 30 |
CFG scale
|
seed |
integer
|
0
Max: 2147483647 |
Seed
|
sd_model |
string
(enum)
|
Stable Diffusion 1.5
Options: Stable Diffusion 1.5, revAnimated_v11, realisticVisionV20_v20 |
Base model
|
a_prompt |
string
|
RAW photo, subject, (high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3
|
Added prompt
|
n_prompt |
string
|
(deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime, mutated hands and fingers:1.4), (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, disconnected limbs, mutation, mutated, ugly, disgusting, amputation
|
Negative prompt
|
interval |
integer
|
1
Min: 1 Max: 10 |
Key frame frequency, K (Uniformly sample the key frames every K frames)
|
keyframe_count |
integer
|
8
Min: 1 Max: 8 |
Total number of key frames (To avoid overload, maximum 8 key frames)
|
x0_strength |
number
|
0.75
Max: 1.05 |
Strength of denoising
|
use_constraints |
string
(enum)
|
shape-aware fusion
Options: shape-aware fusion, pixel-aware fusion, color-aware AdaIN |
Constraints for cross-frame
|
cross_start |
number
|
0
Max: 1 |
Start of cross-frame attention
|
cross_end |
number
|
1
Max: 1 |
End of cross-frame attention
|
style_update_freq |
integer
|
1
Min: 1 Max: 100 |
Frequency of updating for cross-frame attention (Update the key and value for cross-frame attention every N key frames (recommend N*K>=10))
|
warp_start |
number
|
0
Max: 1 |
Start of shape-aware fusion
|
warp_end |
number
|
1
Max: 1 |
End of shape-aware fusion
|
mask_start |
number
|
0.5
Max: 1 |
Start of pixel-aware fusion
|
mask_end |
number
|
0.8
Max: 1 |
End of pixel-aware fusion
|
mask_strength |
number
|
0.5
Max: 1 |
Strength of pixel-aware fusion
|
ada_start |
number
|
0.8
Max: 1 |
Start of color-aware AdaIN
|
ada_end |
number
|
1
Max: 1 |
End of color-aware AdaIN
|
inner_strength |
number
|
0.9
Min: 0.5 Max: 1 |
Pixel-aware fusion detail level (Use a low value to prevent artifacts)
|
smooth_boundary |
boolean
|
True
|
Smooth fusion boundary (Select to prevent artifacts at boundary)
|
Output schema
The shape of the response you’ll get when you run this model with an API.
{'format': 'uri', 'title': 'Output', 'type': 'string'}