{
  "_class_name": "Vid2VidSTUNet",
  "in_channels": 4,
  "out_channels": 4,
  "dim": 320,
  "y_dim": 1024,
  "dim_mult": [
    1,
    2,
    4,
    4
  ],
  "num_heads": 8,
  "head_dim": 64,
  "context_channels": 1024,
  "num_res_blocks": 2,
  "attn_scales": [
    1,
    0.5,
    0.25
  ],
  "dropout": 0.1,
  "temporal_attention": "True",
  "temporal_attn_times": 1,
  "num_tokens": 4,
  "default_fps": 8,
  "input_dim": 1024
}