{
  "architectures": [
    "PerceiverSymbolicAudioModel"
  ],
  "is_decoder": true,
  "model_config": {
    "abs_pos_emb": false,
    "activation_checkpointing": true,
    "activation_offloading": false,
    "cross_attention_dropout": 0.1,
    "cross_attention_widening_factor": 4,
    "init_scale": 0.02,
    "max_heads_parallel": null,
    "max_latents": 2048,
    "max_seq_len": 6144,
    "num_channels": 768,
    "num_heads": 8,
    "num_self_attention_layers": 18,
    "output_bias": false,
    "output_norm": true,
    "post_attention_dropout": 0.1,
    "residual_dropout": 0.1,
    "self_attention_widening_factor": 4,
    "vocab_size": 389
  },
  "model_type": "perceiver-ar-symbolic-audio-model",
  "torch_dtype": "float32",
  "transformers_version": "4.28.0"
}