config.json

{
  "model_type": "wav2vec2-conformer",
  "auto_map": {
    "AutoModel": "modeling_conformer.Wav2Vec2ConformerRNNT"
  },
  "vocab_size": 5632,
  "hidden_size": 512,
  "num_hidden_layers": 17,
  "num_attention_heads": 8,
  "intermediate_size": 2048,
  "hidden_act": "swish",
  "conv_depthwise_kernel_size": 31,
  "mask_time_prob": 0,
  "lstm_layer": 1,
  "pred_hidden": 640,
  "joint_hidden": 640,
  "sampling_rate": 16000,
  "max_symbols_per_step": 10,
  "apply_spec_augment": false,
  "feat_extract_activation": "relu",
  "feat_extract_norm": "layer",
  "conv_bias": true,
  "conv_stride": [2, 2],
  "conv_kernel": [3, 3],
  "conv_dim": [512, 512],
  "blank_id": 256,
  "languages": [
    "as",
    "bn",
    "brx",
    "doi",
    "gu",
    "hi",
    "kn",
    "kok",
    "ks",
    "mai",
    "ml",
    "mni",
    "mr",
    "ne",
    "or",
    "pa",
    "sa",
    "sat",
    "sd",
    "ta",
    "te",
    "ur"
  ]
}