hyperparams.yaml

# ############################################################################
# Model: E2E ASR with attention-based ASR
# Training data: All Finnish
# Encoder: CRDNN
# Decoder: GRU + beamsearch
# Authors: Aku Rouhe 2022
# ############################################################################

num_units: 5000
tokenizer: !new:sentencepiece.SentencePieceProcessor

# Feature parameters
sample_rate: 16000
n_fft: 400
n_mels: 40

# Model parameters
activation: !name:torch.nn.LeakyReLU
dropout: 0.15
cnn_blocks: 2
cnn_channels: (64, 128)
inter_layer_pooling_size: (2, 2)
cnn_kernelsize: (3, 3)
time_pooling_size: 4
rnn_class: !name:speechbrain.nnet.RNN.LSTM
rnn_layers: 3
rnn_neurons: 512
rnn_bidirectional: True
dnn_blocks: 1
dnn_neurons: 512
emb_size: 128
dec_neurons: 1024
dec_layers: 1
output_neurons: !ref <num_units>  

unk_index: 1 
blank_index: 0
pad_index: 0
bos_index: 1
eos_index: 2

min_decode_ratio: 0.0
max_decode_ratio: 1.0
valid_beam_size: 4
test_beam_size: 8
eos_threshold: 1.2
using_max_attn_shift: False
max_attn_shift: 240
ctc_weight_decode: 0.0
coverage_penalty: 3.0 
temperature: 1.5 

# Feature extraction
compute_features: !new:speechbrain.lobes.features.Fbank
    sample_rate: !ref <sample_rate>
    n_fft: !ref <n_fft>
    n_mels: !ref <n_mels>

# Feature normalization (mean and std)
normalize: !new:speechbrain.processing.features.InputNormalization
    norm_type: global
    update_until_epoch: -1

# The CRDNN model is an encoder that combines CNNs, RNNs, and DNNs.
encoder: !new:speechbrain.lobes.models.CRDNN.CRDNN
    input_shape: [null, null, !ref <n_mels>]
    activation: !ref <activation>
    dropout: !ref <dropout>
    cnn_blocks: !ref <cnn_blocks>
    cnn_channels: !ref <cnn_channels>
    cnn_kernelsize: !ref <cnn_kernelsize>
    inter_layer_pooling_size: !ref <inter_layer_pooling_size>
    time_pooling: True
    using_2d_pooling: False
    time_pooling_size: !ref <time_pooling_size>
    rnn_class: !ref <rnn_class>
    rnn_layers: !ref <rnn_layers>
    rnn_neurons: !ref <rnn_neurons>
    rnn_bidirectional: !ref <rnn_bidirectional>
    rnn_re_init: True
    dnn_blocks: !ref <dnn_blocks>
    dnn_neurons: !ref <dnn_neurons>
    use_rnnp: False

# Embedding (from indexes to an embedding space of dimension emb_size).
embedding: !new:speechbrain.nnet.embedding.Embedding
    num_embeddings: !ref <output_neurons>
    embedding_dim: !ref <emb_size>

# Attention-based RNN decoder.
decoder: !new:speechbrain.nnet.RNN.AttentionalRNNDecoder
    enc_dim: !ref <dnn_neurons>
    input_size: !ref <emb_size>
    rnn_type: gru
    attn_type: location
    hidden_size: !ref <dec_neurons>
    attn_dim: 2048
    num_layers: !ref <dec_layers>
    scaling: 1.0
    channels: 10
    kernel_size: 100
    re_init: True
    dropout: !ref <dropout>

# Linear transformation on the top of the encoder.
ctc_lin: !new:speechbrain.nnet.linear.Linear
    input_size: !ref <dnn_neurons>
    n_neurons: !ref <output_neurons>

# Linear transformation on the top of the decoder.
seq_lin: !new:speechbrain.nnet.linear.Linear
    input_size: !ref <dec_neurons>
    n_neurons: !ref <output_neurons>

# Final softmax (for log posteriors computation).
log_softmax: !new:speechbrain.nnet.activations.Softmax
    apply_log: True

# Cost definition for the CTC part.
ctc_cost: !name:speechbrain.nnet.losses.ctc_loss
    blank_index: !ref <blank_index>

full_encode_step: !new:speechbrain.nnet.containers.LengthsCapableSequential
    input_shape: [null, null, !ref <n_mels>]
    compute_features: !ref <compute_features>
    normalize: !ref <normalize>
    model: !ref <encoder>

# Gathering all the submodels in a single model object.
model: !new:torch.nn.ModuleList
    - - !ref <encoder>
      - !ref <embedding>
      - !ref <decoder>
      - !ref <ctc_lin>
      - !ref <seq_lin>

test_search: !new:speechbrain.decoders.S2SRNNBeamSearcher
    embedding: !ref <embedding>
    decoder: !ref <decoder>
    linear: !ref <seq_lin>
    ctc_linear: !ref <ctc_lin>
    bos_index: !ref <bos_index>
    eos_index: !ref <eos_index>
    blank_index: !ref <blank_index>
    min_decode_ratio: !ref <min_decode_ratio>
    max_decode_ratio: !ref <max_decode_ratio>
    beam_size: !ref <test_beam_size>
    eos_threshold: !ref <eos_threshold>
    using_max_attn_shift: !ref <using_max_attn_shift>
    max_attn_shift: !ref <max_attn_shift>
    coverage_penalty: !ref <coverage_penalty>
    ctc_weight: !ref <ctc_weight_decode>
    temperature: !ref <temperature>

# Objects in "modules" dict will have their parameters moved to the correct
# device, as well as having train()/eval() called on them by the Brain class
modules:
    encoder: !ref <full_encode_step>
    decoder: !ref <test_search>

pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
    loadables:
        model: !ref <model>
        normalizer: !ref <normalize>
        tokenizer: !ref <tokenizer>