# network architecture
|
# encoder related
|
encoder: transformer
|
encoder_conf:
|
output_size: 256 # dimension of attention
|
attention_heads: 4
|
linear_units: 2048 # the number of units of position-wise feed forward
|
num_blocks: 12 # the number of encoder blocks
|
dropout_rate: 0.1
|
positional_dropout_rate: 0.1
|
attention_dropout_rate: 0.0
|
input_layer: conv2d # encoder architecture type
|
normalize_before: true
|
|
# decoder related
|
decoder: transformer
|
decoder_conf:
|
attention_heads: 4
|
linear_units: 2048
|
num_blocks: 6
|
dropout_rate: 0.1
|
positional_dropout_rate: 0.1
|
self_attention_dropout_rate: 0.0
|
src_attention_dropout_rate: 0.0
|
|
# frontend related
|
frontend: wav_frontend
|
frontend_conf:
|
fs: 16000
|
window: hamming
|
n_mels: 80
|
frame_length: 25
|
frame_shift: 10
|
lfr_m: 1
|
lfr_n: 1
|
|
# hybrid CTC/attention
|
model_conf:
|
ctc_weight: 0.3
|
lsm_weight: 0.1 # label smoothing option
|
length_normalized_loss: false
|
|
# optimization related
|
accum_grad: 1
|
grad_clip: 5
|
patience: none
|
max_epoch: 60
|
val_scheduler_criterion:
|
- valid
|
- acc
|
best_model_criterion:
|
- - valid
|
- acc
|
- max
|
keep_nbest_models: 10
|
|
# NoamLR is deprecated. Use WarmupLR.
|
# The following is equivalent setting for NoamLR:
|
#
|
# optim: adam
|
# optim_conf:
|
# lr: 10.
|
# scheduler: noamlr
|
# scheduler_conf:
|
# model_size: 256
|
# warmup_steps: 25000
|
#
|
optim: adam
|
optim_conf:
|
lr: 0.002
|
scheduler: warmuplr # pytorch v1.1.0+ required
|
scheduler_conf:
|
warmup_steps: 25000
|
|
dataset_conf:
|
data_names: speech,text
|
data_types: sound,text
|
shuffle: True
|
shuffle_conf:
|
shuffle_size: 2048
|
sort_size: 500
|
batch_conf:
|
batch_type: token
|
batch_size: 25000
|
num_workers: 8
|
|
log_interval: 50
|
normalize: None
|
