{
  "architectures": [
    "VASTDiT3"
  ],
  "audio_hidden_size": 1152,
  "audio_in_channels": 8,
  "audio_input_size": [
    null,
    null
  ],
  "audio_patch_size": [
    4,
    1
  ],
  "caption_channels": 4096,
  "class_dropout_prob": 0.1,
  "depth": 28,
  "drop_path": 0.0,
  "enable_flash_attn": true,
  "enable_layernorm_kernel": false,
  "enable_sequence_parallelism": false,
  "freeze_audio_branch": false,
  "freeze_video_branch": true,
  "freeze_y_embedder": false,
  "hidden_size": 1152,
  "in_channels": 4,
  "input_size": [
    null,
    null,
    null
  ],
  "input_sq_size": 512,
  "mlp_ratio": 4.0,
  "model_max_length": 300,
  "model_type": "VASTDiT3",
  "num_heads": 16,
  "only_infer_audio": false,
  "only_train_audio": true,
  "only_train_temporal": false,
  "patch_size": [
    1,
    2,
    2
  ],
  "pred_sigma": true,
  "qk_norm": true,
  "skip_y_embedder": false,
  "spatial_prior_len": 16,
  "st_prior_channel": 64,
  "temporal_prior_len": 16,
  "torch_dtype": "bfloat16",
  "train_st_prior_attn": false,
  "train_va_cross_attn": false,
  "transformers_version": "4.39.3",
  "weight_init_from": [
    "./checkpoints/OpenSora-STDiT-v3/model.safetensors"
  ]
}