{
  "activation_dropout": 0.0,
  "adapter_act": "relu",
  "adapter_kernel_size": 3,
  "adapter_stride": 2,
  "add_adapter": false,
  "apply_spec_augment": true,
  "architectures": [
    "Wav2Vec2BertForSequenceClassification"
  ],
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "classifier_proj_size": 768,
  "conformer_conv_dropout": 0.1,
  "conv_depthwise_kernel_size": 31,
  "ctc_loss_reduction": "sum",
  "ctc_zero_infinity": false,
  "do_stable_layer_norm": false,
  "dtype": "float32",
  "eos_token_id": 2,
  "feat_proj_dropout": 0.0,
  "feature_projection_input_dim": 10,
  "final_dropout": 0.1,
  "hidden_act": "gelu",
  "hidden_dropout": 0.0,
  "hidden_dropout_prob": 0.0,
  "hidden_size": 120,
  "initializer_range": 0.02,
  "intermediate_size": 216,
  "layer_norm_eps": 1e-05,
  "layerdrop": 0.1,
  "left_max_position_embeddings": 64,
  "mask_feature_length": 10,
  "mask_feature_min_masks": 0,
  "mask_feature_prob": 0.0,
  "mask_time_length": 2,
  "mask_time_min_masks": 2,
  "mask_time_prob": 0.0,
  "max_source_positions": 5000,
  "model_type": "wav2vec2-bert",
  "num_adapter_layers": 2,
  "num_attention_heads": 2,
  "num_conv_pos_embedding_groups": 2,
  "num_conv_pos_embeddings": 16,
  "num_hidden_layers": 4,
  "output_hidden_size": 120,
  "pad_token_id": 0,
  "position_embeddings_type": "relative",
  "right_max_position_embeddings": 8,
  "rotary_embedding_base": 10000,
  "tdnn_dilation": [
    1,
    2
  ],
  "tdnn_dim": [
    4,
    4
  ],
  "tdnn_kernel": [
    3,
    3
  ],
  "transformers_version": "5.16.0.dev0",
  "use_intermediate_ffn_before_adapter": false,
  "use_weighted_layer_sum": false,
  "vocab_size": 16,
  "xvector_output_dim": 2
}