{
  "_name_or_path": "MIT/ast-finetuned-speech-commands-v2",
  "architectures": [
    "ASTForAudioClassification"
  ],
  "attention_probs_dropout_prob": 0.0,
  "frequency_stride": 10,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "id2label": {
    "0": "yes",
    "1": "no",
    "10": "zero",
    "11": "one",
    "12": "two",
    "13": "three",
    "14": "four",
    "15": "five",
    "16": "six",
    "17": "seven",
    "18": "eight",
    "19": "nine",
    "2": "up",
    "20": "bed",
    "21": "bird",
    "22": "cat",
    "23": "dog",
    "24": "happy",
    "25": "house",
    "26": "marvin",
    "27": "sheila",
    "28": "tree",
    "29": "wow",
    "3": "down",
    "30": "backward",
    "31": "forward",
    "32": "follow",
    "33": "learn",
    "34": "visual",
    "35": "_silence_",
    "36": "player2",
    "4": "left",
    "5": "right",
    "6": "on",
    "7": "off",
    "8": "stop",
    "9": "go"
  },
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "_silence_": "35",
    "backward": "30",
    "bed": "20",
    "bird": "21",
    "cat": "22",
    "dog": "23",
    "down": "3",
    "eight": "18",
    "five": "15",
    "follow": "32",
    "forward": "31",
    "four": "14",
    "go": "9",
    "happy": "24",
    "house": "25",
    "learn": "33",
    "left": "4",
    "marvin": "26",
    "nine": "19",
    "no": "1",
    "off": "7",
    "on": "6",
    "one": "11",
    "player2": "36",
    "right": "5",
    "seven": "17",
    "sheila": "27",
    "six": "16",
    "stop": "8",
    "three": "13",
    "tree": "28",
    "two": "12",
    "up": "2",
    "visual": "34",
    "wow": "29",
    "yes": "0",
    "zero": "10"
  },
  "layer_norm_eps": 1e-12,
  "max_length": 128,
  "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": "4.38.2"
}