config.json

{
  "_name_or_path": "ajesujoba/AfriHuBERT",
  "activation_dropout": 0.1,
  "apply_spec_augment": true,
  "architectures": [
    "HubertForSequenceClassification"
  ],
  "attention_dropout": 0.1,
  "bos_token_id": 1,
  "classifier_proj_size": 256,
  "conv_bias": false,
  "conv_dim": [
    512,
    512,
    512,
    512,
    512,
    512,
    512
  ],
  "conv_kernel": [
    10,
    3,
    3,
    3,
    3,
    2,
    2
  ],
  "conv_pos_batch_norm": false,
  "conv_stride": [
    5,
    2,
    2,
    2,
    2,
    2,
    2
  ],
  "ctc_loss_reduction": "sum",
  "ctc_zero_infinity": false,
  "do_stable_layer_norm": false,
  "eos_token_id": 2,
  "feat_extract_activation": "gelu",
  "feat_extract_dropout": 0.0,
  "feat_extract_norm": "group",
  "feat_proj_dropout": 0.1,
  "feat_proj_layer_norm": true,
  "final_dropout": 0.1,
  "finetuning_task": "audio-classification",
  "gradient_checkpointing": false,
  "hidden_act": "gelu",
  "hidden_dropout": 0.1,
  "hidden_dropout_prob": 0.1,
  "hidden_size": 768,
  "id2label": {
    "0": "Akuapim-twi",
    "1": "Asante-twi",
    "10": "eng",
    "11": "eng-zul",
    "12": "ewe",
    "13": "fat",
    "14": "fon",
    "15": "fuc",
    "16": "fuf",
    "17": "gaa",
    "18": "hau",
    "19": "ibo",
    "2": "aeb",
    "20": "kab",
    "21": "kin",
    "22": "kln",
    "23": "lin",
    "24": "loz",
    "25": "lug",
    "26": "luo",
    "27": "mlq",
    "28": "nbl",
    "29": "nso",
    "3": "afr",
    "30": "nya",
    "31": "sot",
    "32": "srr",
    "33": "ssw",
    "34": "sus",
    "35": "swa",
    "36": "swh",
    "37": "tig",
    "38": "tir",
    "39": "toi",
    "4": "amh",
    "40": "tsn",
    "41": "tso",
    "42": "twi",
    "43": "ven",
    "44": "wol",
    "45": "xho",
    "46": "yor",
    "47": "zgh",
    "48": "zul",
    "5": "ara",
    "6": "bas",
    "7": "bem",
    "8": "dav",
    "9": "dyu"
  },
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "Akuapim-twi": "0",
    "Asante-twi": "1",
    "aeb": "2",
    "afr": "3",
    "amh": "4",
    "ara": "5",
    "bas": "6",
    "bem": "7",
    "dav": "8",
    "dyu": "9",
    "eng": "10",
    "eng-zul": "11",
    "ewe": "12",
    "fat": "13",
    "fon": "14",
    "fuc": "15",
    "fuf": "16",
    "gaa": "17",
    "hau": "18",
    "ibo": "19",
    "kab": "20",
    "kin": "21",
    "kln": "22",
    "lin": "23",
    "loz": "24",
    "lug": "25",
    "luo": "26",
    "mlq": "27",
    "nbl": "28",
    "nso": "29",
    "nya": "30",
    "sot": "31",
    "srr": "32",
    "ssw": "33",
    "sus": "34",
    "swa": "35",
    "swh": "36",
    "tig": "37",
    "tir": "38",
    "toi": "39",
    "tsn": "40",
    "tso": "41",
    "twi": "42",
    "ven": "43",
    "wol": "44",
    "xho": "45",
    "yor": "46",
    "zgh": "47",
    "zul": "48"
  },
  "layer_norm_eps": 1e-05,
  "layerdrop": 0.1,
  "mask_feature_length": 10,
  "mask_feature_min_masks": 0,
  "mask_feature_prob": 0.0,
  "mask_time_length": 10,
  "mask_time_min_masks": 2,
  "mask_time_prob": 0.05,
  "model_type": "hubert",
  "num_attention_heads": 12,
  "num_conv_pos_embedding_groups": 16,
  "num_conv_pos_embeddings": 128,
  "num_feat_extract_layers": 7,
  "num_hidden_layers": 12,
  "pad_token_id": 0,
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
  "torch_dtype": "float32",
  "transformers_version": "4.48.1",
  "use_weighted_layer_sum": false,
  "vocab_size": 32
}