forked from speechbrain/speechbrain
-
Notifications
You must be signed in to change notification settings - Fork 0
/
train_ecapa.yaml
170 lines (139 loc) · 5.34 KB
/
train_ecapa.yaml
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
################################
# Model: language identification with ECAPA
# Authors: Tanel Alumäe, 2021
# ################################
# Basic parameters
seed: 1988
__set_seed: !apply:torch.manual_seed [!ref <seed>]
output_folder: !ref results/epaca/<seed>
save_folder: !ref <output_folder>/save
train_log: !ref <output_folder>/train_log.txt
data_folder: !PLACEHOLDER
shards_url: /data/voxlingua107_shards
train_meta: !ref <shards_url>/train/meta.json
val_meta: !ref <shards_url>/dev/meta.json
train_shards: !ref <shards_url>/train/shard-{000000..000507}.tar
val_shards: !ref <shards_url>/dev/shard-000000.tar
# Data for augmentation
NOISE_DATASET_URL: https://www.dropbox.com/scl/fi/a09pj97s5ifan81dqhi4n/noises.zip?rlkey=j8b0n9kdjdr32o1f06t0cw5b7&dl=1
RIR_DATASET_URL: https://www.dropbox.com/scl/fi/linhy77c36mu10965a836/RIRs.zip?rlkey=pg9cu8vrpn2u173vhiqyu743u&dl=1
data_folder_noise: !ref <data_folder>/noise # The noisy sequences for data augmentation will automatically be downloaded here.
data_folder_rir: !ref <data_folder>/rir # The impulse responses used for data augmentation will automatically be downloaded here.
noise_annotation: !ref <save_folder>/noise.csv
rir_annotation: !ref <save_folder>/rir.csv
# Set to directory on a large disk if you are training on Webdataset shards hosted on the web
shard_cache_dir:
ckpt_interval_minutes: 5
# Training parameters
number_of_epochs: 40
lr: 0.001
lr_final: 0.0001
sample_rate: 16000
sentence_len: 3 # seconds
# Feature parameters
n_mels: 60
left_frames: 0
right_frames: 0
deltas: False
# Number of languages
out_n_neurons: 107
num_workers: 4
batch_size: 128
batch_size_val: 32
train_dataloader_options:
num_workers: !ref <num_workers>
batch_size: !ref <batch_size>
val_dataloader_options:
num_workers: 1
batch_size: !ref <batch_size_val>
############################## Augmentations ###################################
# Download and prepare the dataset of noisy sequences for augmentation
prepare_noise_data: !name:speechbrain.augment.preparation.prepare_dataset_from_URL
URL: !ref <NOISE_DATASET_URL>
dest_folder: !ref <data_folder_noise>
ext: wav
csv_file: !ref <noise_annotation>
# Download and prepare the dataset of room impulse responses for augmentation
prepare_rir_data: !name:speechbrain.augment.preparation.prepare_dataset_from_URL
URL: !ref <RIR_DATASET_URL>
dest_folder: !ref <data_folder_rir>
ext: wav
csv_file: !ref <rir_annotation>
# Add reverberation to input signal
add_reverb: !new:speechbrain.augment.time_domain.AddReverb
csv_file: !ref <rir_annotation>
reverb_sample_rate: !ref <sample_rate>
clean_sample_rate: !ref <sample_rate>
num_workers: !ref <num_workers>
# Add noise to input signal
add_noise: !new:speechbrain.augment.time_domain.AddNoise
csv_file: !ref <noise_annotation>
snr_low: 0
snr_high: 15
noise_sample_rate: !ref <sample_rate>
clean_sample_rate: !ref <sample_rate>
num_workers: !ref <num_workers>
# Speed perturbation
speed_perturb: !new:speechbrain.augment.time_domain.SpeedPerturb
orig_freq: !ref <sample_rate>
# Augmenter: Combines previously defined augmentations to perform data augmentation
wav_augment: !new:speechbrain.augment.augmenter.Augmenter
concat_original: True
shuffle_augmentations: True
min_augmentations: 1
max_augmentations: 3
augmentations: [
!ref <add_reverb>,
!ref <add_noise>,
!ref <speed_perturb>]
# Functions
compute_features: !new:speechbrain.lobes.features.Fbank
n_mels: !ref <n_mels>
left_frames: !ref <left_frames>
right_frames: !ref <right_frames>
deltas: !ref <deltas>
embedding_model: !new:speechbrain.lobes.models.ECAPA_TDNN.ECAPA_TDNN
input_size: !ref <n_mels>
channels: [1024, 1024, 1024, 1024, 3072]
kernel_sizes: [5, 3, 3, 3, 1]
dilations: [1, 2, 3, 4, 1]
attention_channels: 128
lin_neurons: 256
classifier: !new:speechbrain.lobes.models.Xvector.Classifier
input_shape: [null, null, 256]
activation: !name:torch.nn.LeakyReLU
lin_blocks: 1
lin_neurons: 512
out_neurons: !ref <out_n_neurons>
epoch_counter: !new:speechbrain.utils.epoch_loop.EpochCounter
limit: !ref <number_of_epochs>
mean_var_norm: !new:speechbrain.processing.features.InputNormalization
norm_type: sentence
std_norm: False
modules:
compute_features: !ref <compute_features>
embedding_model: !ref <embedding_model>
classifier: !ref <classifier>
mean_var_norm: !ref <mean_var_norm>
compute_cost: !name:speechbrain.nnet.losses.nll_loss
# compute_error: !name:speechbrain.nnet.losses.classification_error
opt_class: !name:torch.optim.Adam
lr: !ref <lr>
weight_decay: 0.000002
lr_annealing: !new:speechbrain.nnet.schedulers.LinearScheduler
initial_value: !ref <lr>
final_value: !ref <lr_final>
epoch_count: !ref <number_of_epochs>
# Logging + checkpoints
train_logger: !new:speechbrain.utils.train_logger.FileTrainLogger
save_file: !ref <train_log>
error_stats: !name:speechbrain.utils.metric_stats.MetricStats
metric: !name:speechbrain.nnet.losses.classification_error
reduction: batch
checkpointer: !new:speechbrain.utils.checkpoints.Checkpointer
checkpoints_dir: !ref <save_folder>
recoverables:
embedding_model: !ref <embedding_model>
classifier: !ref <classifier>
normalizer: !ref <mean_var_norm>
counter: !ref <epoch_counter>