pretrained_model_path: checkpoints/stable-diffusion-v1-4
output_dir: outputs/child-riding

input_data:
  video_path: data/child-riding.mp4
  prompt: "a child is riding a bike on the road"
  n_sample_frames: 8
  width: 512
  height: 512
  sample_start_idx: 0
  sample_frame_rate: 1

validation_data:
  # inv_latent: "outputs_2d/car-turn/inv_latents/ddim_latent-0.pt"  # latent inversed w/o SCAttn ! 
  prompts:
    - a lego child is riding a bike on the road
    - a child is riding a bike on the flooded road
  video_length: 8
  width: 512
  height: 512
  num_inference_steps: 50
  guidance_scale: 7.5
  num_inv_steps: 50
  # args for null-text inv
  use_null_inv: True
  null_inner_steps: 1
  null_base_lr: 1e-2
  null_uncond_ratio: -0.5
  null_normal_infer: True

input_batch_size: 1
seed: 33
mixed_precision: "no"
gradient_checkpointing: True
enable_xformers_memory_efficient_attention: True
# test-time adaptation
use_sc_attn: True
use_st_attn: True
st_attn_idx: 0