{ "model_type": "AudioVideo", "num_attention_heads": 32, "attention_head_dim": 128, "in_channels": 128, "out_channels": 128, "num_layers": 48, "cross_attention_dim": 4096, "caption_channels": null, "audio_num_attention_heads": 32, "audio_attention_head_dim": 64, "audio_in_channels": 128, "audio_out_channels": 128, "audio_cross_attention_dim": 2048, "positional_embedding_theta": 10000.0, "positional_embedding_max_pos": [ 20, 2048, 2048 ], "audio_positional_embedding_max_pos": [ 20 ], "timestep_scale_multiplier": 1000, "av_ca_timestep_scale_multiplier": 1000, "norm_eps": 1e-06, "audio_sample_rate": 24000, "audio_latent_sample_rate": 16000, "audio_hop_length": 160, "audio_latent_channels": 8, "audio_mel_bins": 16, "model_version": "2.3.0", "is_v2": true, "apply_gated_attention": true, "cross_attention_adaln": true, "connector_positional_embedding_max_pos": [ 4096 ], "connector_rope_type": "SPLIT", "connector_num_attention_heads": 32, "connector_attention_head_dim": 128, "audio_connector_num_attention_heads": 32, "audio_connector_attention_head_dim": 64, "caption_projection_first_linear": false, "caption_projection_second_linear": false }