hyperparams.yaml

# ############################################################################
# Model: CTC-wav2vec2
# Encoder: wav2vec2
# Decoder: -
# Tokens: Char
# losses: CTC
# Training: AISHELL-1
# Authors:  Yingzhi WANG 2022
# ############################################################################

wav2vec2_hub: TencentGameMate/chinese-wav2vec2-large
sample_rate: 16000

wav2vec_output_dim: 1024
dnn_neurons: 1024
freeze_wav2vec: False
dropout: 0.15

tokenizer: !apply:transformers.BertTokenizer.from_pretrained
   pretrained_model_name_or_path: bert-base-chinese

output_neurons: 21128

# Decoding parameters
# Be sure that the bos and eos index match with the BPEs ones
blank_index: 0

enc: !new:speechbrain.nnet.containers.Sequential
   input_shape: [null, null, !ref <wav2vec_output_dim>]
   linear1: !name:speechbrain.nnet.linear.Linear
      n_neurons: !ref <dnn_neurons>
      bias: True
   bn1: !name:speechbrain.nnet.normalization.BatchNorm1d
   activation: !new:torch.nn.LeakyReLU
   drop: !new:torch.nn.Dropout
      p: !ref <dropout>
   linear2: !name:speechbrain.nnet.linear.Linear
      n_neurons: !ref <dnn_neurons>
      bias: True
   bn2: !name:speechbrain.nnet.normalization.BatchNorm1d
   activation2: !new:torch.nn.LeakyReLU
   drop2: !new:torch.nn.Dropout
      p: !ref <dropout>
   linear3: !name:speechbrain.nnet.linear.Linear
      n_neurons: !ref <dnn_neurons>
      bias: True
   bn3: !name:speechbrain.nnet.normalization.BatchNorm1d
   activation3: !new:torch.nn.LeakyReLU

wav2vec2: !new:speechbrain.lobes.models.huggingface_transformers.wav2vec2.Wav2Vec2
   source: !ref <wav2vec2_hub>
   output_norm: True
   freeze: !ref <freeze_wav2vec>
   save_path: model_checkpoints

ctc_lin: !new:speechbrain.nnet.linear.Linear
   input_size: !ref <dnn_neurons>
   n_neurons: !ref <output_neurons>

log_softmax: !new:speechbrain.nnet.activations.Softmax
   apply_log: True

decoder: !name:speechbrain.decoders.ctc_greedy_decode
   blank_id: !ref <blank_index>

model: !new:torch.nn.ModuleList
   - [!ref <enc>, !ref <ctc_lin>]

modules:
   wav2vec2: !ref <wav2vec2>
   enc: !ref <enc>
   ctc_lin: !ref <ctc_lin>
   log_softmax: !ref <log_softmax>

pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
   loadables:
      wav2vec2: !ref <wav2vec2>
      model: !ref <model>