ComfyUI Node: Animate Anyone Sampler

Authored by MrForExample

Created 9 months ago

Updated 5 months ago

489 stars

Category

AnimateAnyone-Evolved

Inputs

reference_unet UNET2D

denoising_unet UNET3D

ref_image_latent LATENT

clip_image_embeds CLIP_VISION_OUTPUT

pose_latent POSE_LATENT

seed INT

steps INT

cfg FLOAT

delta FLOAT

context_frames INT

context_stride INT

context_overlap INT

context_batch_size INT

interpolation_factor INT

sampler_scheduler_pairs

• DDIM
• DPM++ 2M Karras
• LCM
• Euler
• Euler Ancestral
• LMS
• PNDM

beta_start FLOAT

beta_end FLOAT

beta_schedule

• linear
• scaled_linear
• squaredcos_cap_v2

prediction_type

• v_prediction
• epsilon
• sample

timestep_spacing

• trailing
• linspace
• leading

steps_offset INT

clip_sample BOOLEAN

rescale_betas_zero_snr BOOLEAN

use_lora BOOLEAN

lora_name

Outputs

LATENT

Extension: ComfyUI-AnimateAnyone-Evolved

Improved AnimateAnyone implementation that allows you to use the opse image sequence and reference image to generate stylized video. The current goal of this project is to achieve desired pose2video result with 1+FPS on GPUs that are equal to or better than RTX 3080!🚀 [w/The torch environment may be compromised due to version issues as some torch-related packages are being reinstalled.]

Authored by MrForExample