adnovamedya/deforum-sd

stable diffusion deforum model

Public
406 runs

Run adnovamedya/deforum-sd with an API

Use one of our client libraries to get started quickly. Clicking on a library will take you to the Playground tab where you can tweak different inputs, see the results, and copy the corresponding code to use in your own project.

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
model_checkpoint
None
Protogen_V2.2.ckpt
Choose stable diffusion model.
max_frames
integer
200
Number of frames for animation
animation_prompts
string
0: a beautiful apple, trending on Artstation | 50: a beautiful banana, trending on Artstation | 100: a beautiful coconut, trending on Artstation | 150: a beautiful durian, trending on Artstation
Prompt for animation. Provide 'frame number : prompt at this frame', separate different prompts with '|'. Make sure the frame number does not exceed the max_frames.
negative_prompts
string
0: mountain
Prompt for negative. Provide 'frame number : prompt at this frame', separate different prompts with '|'. Make sure the frame number does not exceed the max_frames.
width
None
360
Width of output video. Reduce if out of memory.
height
None
640
Height of output image. Reduce if out of memory.
num_inference_steps
integer
50

Min: 1

Max: 500

Number of denoising steps
guidance_scale
number
7

Min: 1

Max: 20

Scale for classifier-free guidance
sampler
None
euler_ancestral
None
seed
integer
Random seed. Leave blank to randomize the seed
fps
integer
15

Min: 10

Max: 60

Choose fps for the video.
clip_name
None
ViT-L/14
Choose CLIP model
use_init
boolean
False
If not using init image, you can skip the next settings to setting the animation_mode.
init_image
string
Provide init_image if use_init
add_init_noise
boolean
False
add_init_noise
strength
number
0.5
The initial diffusion on the input image.
use_mask
boolean
False
None
mask_file
string
Provide mask_file if use_mask
invert_mask
boolean
False
None
animation_mode
None
2D
Choose Animation mode. All parameters below are for setting up animations.
border
None
replicate
None
angle
string
0:(0)
angle parameter for the motion
zoom
string
0:(1.04)
zoom parameter for the motion
translation_x
string
0:(10*sin(2*3.14*t/10))
translation_x parameter for the 2D motion
translation_y
string
0:(0)
translation_y parameter for the 2D motion
translation_z
string
0:(10)
translation_z parameter for the 2D motion
rotation_3d_x
string
0:(0)
rotation_3d_x parameter for the 3D motion
rotation_3d_y
string
0:(0)
rotation_3d_y parameter for the 3D motion
rotation_3d_z
string
0:(0)
rotation_3d_z parameter for the 3D motion
flip_2d_perspective
boolean
False
None
perspective_flip_theta
string
0:(0)
None
perspective_flip_phi
string
0:(t%15)
None
perspective_flip_gamma
string
0:(0)
None
perspective_flip_fv
string
0:(53)
None
noise_schedule
string
0: (0.02)
None
strength_schedule
string
0: (0.65)
None
contrast_schedule
string
0: (1.0)
None
hybrid_video_comp_alpha_schedule
string
0:(1)
None
hybrid_video_comp_mask_blend_alpha_schedule
string
0:(0.5)
None
hybrid_video_comp_mask_contrast_schedule
string
0:(1)
None
hybrid_video_comp_mask_auto_contrast_cutoff_high_schedule
string
0:(100)
None
hybrid_video_comp_mask_auto_contrast_cutoff_low_schedule
string
0:(0)
None
enable_schedule_samplers
boolean
False
None
sampler_schedule
string
0:('euler'),10:('dpm2'),20:('dpm2_ancestral'),30:('heun'),40:('euler'),50:('euler_ancestral'),60:('dpm_fast'),70:('dpm_adaptive'),80:('dpmpp_2s_a'),90:('dpmpp_2m')
None
kernel_schedule
string
0: (5)
None
sigma_schedule
string
0: (1.0)
None
amount_schedule
string
0: (0.2)
None
threshold_schedule
string
0: (0.0)
None
color_coherence
None
Match Frame 0 LAB
None
color_coherence_video_every_N_frames
integer
1
None
color_force_grayscale
boolean
False
None
diffusion_cadence
None
1
None
use_depth_warping
boolean
True
None
midas_weight
number
0.3
None
near_plane
integer
200
None
far_plane
integer
10000
None
fov
integer
40
None
padding_mode
None
border
None
sampling_mode
None
bicubic
None
video_init_path
string
None
extract_nth_frame
integer
1
None
overwrite_extracted_frames
boolean
True
None
use_mask_video
boolean
False
None
video_mask_path
string
None
hybrid_generate_inputframes
boolean
False
None
hybrid_use_first_frame_as_init_image
boolean
True
None
hybrid_motion
None
None
None
hybrid_flow_method
None
Farneback
None
hybrid_composite
boolean
False
None
hybrid_comp_mask_type
None
None
None
hybrid_comp_mask_inverse
boolean
False
None
hybrid_comp_mask_equalize
None
None
None
hybrid_comp_mask_auto_contrast
boolean
False
None
hybrid_comp_save_extra_frames
boolean
False
None
hybrid_use_video_as_mse_image
boolean
False
None
interpolate_key_frames
boolean
False
None
interpolate_x_frames
integer
4
None
resume_from_timestring
boolean
False
None
resume_timestring
string
None

Output schema

The shape of the response you’ll get when you run this model with an API.

Schema
{
  "type": "string",
  "title": "Output",
  "format": "uri"
}