File size: 6,117 Bytes
f4950b1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
name: &name "QuartzNet15x5"

model:
  sample_rate: &sample_rate 16000
  repeat: &repeat 5
  dropout: &dropout 0.0
  separable: &separable true
  labels: &labels [" ", "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z", "á", "é", "í", "ñ", "ó", "ú", "ü"]

  train_ds:
    manifest_filepath: ???
    sample_rate: 16000
    labels: *labels
    batch_size: 16  ##########################
    trim_silence: True
    max_duration: 16.7
    shuffle: True
    num_workers: 8
    pin_memory: true
    # tarred datasets
    is_tarred: false
    tarred_audio_filepaths: null
    shuffle_n: 2048
    # bucketing params
    bucketing_strategy: "synced_randomized"
    bucketing_batch_size: null

  validation_ds:
    manifest_filepath: ???
    sample_rate: 16000
    labels: *labels
    batch_size: 16  ##########################
    shuffle: False
    num_workers: 8
    pin_memory: true

  preprocessor:
    _target_: nemo.collections.asr.modules.AudioToMelSpectrogramPreprocessor
    normalize: "per_feature"
    window_size: 0.02
    sample_rate: *sample_rate
    window_stride: 0.01
    window: "hann"
    features: &n_mels 64
    n_fft: 512
    frame_splicing: 1
    dither: 1.0e-05

  spec_augment:
    _target_: nemo.collections.asr.modules.SpectrogramAugmentation
    rect_freq: 50
    rect_masks: 5
    rect_time: 120

  encoder:
    _target_: nemo.collections.asr.modules.ConvASREncoder
    feat_in: *n_mels
    activation: relu
    conv_mask: true

    jasper:
    #1
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [33]
      repeat: 1
      residual: false
      separable: *separable
      stride: [2]
    #2
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [33]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #3
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [33]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #4
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [33]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #5
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [39]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #6
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [39]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #7
    - dilation: [1]
      dropout: *dropout
      filters: 256
      kernel: [39]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #8
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [51]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #9
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [51]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #10
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [51]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #11
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [63]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #12
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [63]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #13
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [63]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #14
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [75]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #15
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [75]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #16
    - dilation: [1]
      dropout: *dropout
      filters: 512
      kernel: [75]
      repeat: *repeat
      residual: true
      separable: *separable
      stride: [1]
    #17
    - dilation: [2]
      dropout: *dropout
      filters: 512
      kernel: [87]
      repeat: 1
      residual: false
      separable: *separable
      stride: [1]
    #18
    - dilation: [1]
      dropout: *dropout
      filters: &enc_filters 1024
      kernel: [1]
      repeat: 1
      residual: false
      stride: [1]

  decoder:
    _target_: nemo.collections.asr.modules.ConvASRDecoder
    feat_in: *enc_filters
    num_classes: 34
    vocabulary: *labels

  optim:
    name: novograd
    # _target_: nemo.core.optim.optimizers.Novograd
    lr: 0.0012
    # optimizer arguments
    betas: [0.8, 0.5]
    weight_decay: 0.001

    # scheduler setup
    sched:
      name: CosineAnnealing

      # pytorch lightning args
      # monitor: val_loss
      # reduce_on_plateau: false

      # Scheduler params
      warmup_steps: null
      warmup_ratio: null
      min_lr: 0.0
      last_epoch: -1

trainer:
  devices: 1 # number of gpus
  max_epochs: 5
  max_steps: -1 # computed at runtime if not set
  num_nodes: 1
  accelerator: gpu
  strategy: ddp
  accumulate_grad_batches: 1
  enable_checkpointing: False  # Provided by exp_manager
  logger: False  # Provided by exp_manager
  log_every_n_steps: 1  # Interval of logging.
  val_check_interval: 1.0  # Set to 0.25 to check 4 times per epoch, or an int for number of iterations
  benchmark: false # needs to be false for models with variable-length speech input as it slows down training

exp_manager:
  exp_dir: null
  name: *name
  create_tensorboard_logger: True
  create_checkpoint_callback: True
  checkpoint_callback_params:
    monitor: "val_wer"
    mode: "min"
  create_wandb_logger: False
  wandb_logger_kwargs:
    name: null
    project: null