{
  "_name_or_path": "MIT/ast-finetuned-audioset-10-10-0.4593",
  "architectures": [
    "ASTForAudioClassification"
  ],
  "attention_probs_dropout_prob": 0.0,
  "frequency_stride": 10,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "id2label": {
    "0": "airport",
    "1": "bus",
    "2": "metro",
    "3": "metro_station",
    "4": "park",
    "5": "public_square",
    "6": "shopping_mall",
    "7": "street_pedestrian",
    "8": "street_traffic",
    "9": "tram"
  },
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "airport": 0,
    "bus": 1,
    "metro": 2,
    "metro_station": 3,
    "park": 4,
    "public_square": 5,
    "shopping_mall": 6,
    "street_pedestrian": 7,
    "street_traffic": 8,
    "tram": 9
  },
  "layer_norm_eps": 1e-12,
  "max_length": 1024,
  "model_type": "audio-spectrogram-transformer",
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "num_mel_bins": 128,
  "patch_size": 16,
  "problem_type": "single_label_classification",
  "qkv_bias": true,
  "time_stride": 10,
  "torch_dtype": "float32",
  "transformers_version": "4.33.2"
}