{
  "_class_name": "WanS2VTransformer3DModel",
  "_diffusers_version": "0.36.0.dev0",
  "adain_mode": "attn_norm",
  "add_last_motion": true,
  "added_kv_proj_dim": null,
  "attention_head_dim": 128,
  "audio_dim": 1024,
  "audio_inject_layers": [
    0,
    4,
    8,
    12,
    16,
    20,
    24,
    27,
    30,
    33,
    36,
    39
  ],
  "cross_attn_norm": true,
  "enable_adain": true,
  "enable_framepack": true,
  "eps": 1e-06,
  "ffn_dim": 13824,
  "framepack_drop_mode": "padd",
  "freq_dim": 256,
  "in_channels": 16,
  "num_attention_heads": 40,
  "num_layers": 40,
  "out_channels": 16,
  "patch_size": [
    1,
    2,
    2
  ],
  "pose_dim": 16,
  "qk_norm": "rms_norm_across_heads",
  "rope_max_seq_len": 1024,
  "text_dim": 4096,
  "zero_timestep": true
}