|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
seed: 1988 |
|
__set_seed: !apply:torch.manual_seed [1988] |
|
output_folder: results_3lang/epaca/1988 |
|
save_folder: results_3lang/epaca/1988/save |
|
train_log: results_3lang/epaca/1988/train_log.txt |
|
data_folder: ./ |
|
rir_folder: ./ |
|
|
|
shards_url: /opt/acoustic-pr/speechbrain_Voxlingua/data_shards |
|
train_meta: /opt/acoustic-pr/speechbrain_Voxlingua/data_shards/train/meta.json |
|
val_meta: /opt/acoustic-pr/speechbrain_Voxlingua/data_shards/dev/meta.json |
|
train_shards: /opt/acoustic-pr/speechbrain_Voxlingua/data_shards/train/shard-{000000..000013}.tar |
|
val_shards: /opt/acoustic-pr/speechbrain_Voxlingua/data_shards/dev/shard-000000.tar |
|
|
|
|
|
|
|
|
|
ckpt_interval_minutes: 5 |
|
|
|
|
|
number_of_epochs: 40 |
|
lr: 0.001 |
|
lr_final: 0.0001 |
|
sample_rate: 16000 |
|
sentence_len: 3 |
|
|
|
|
|
n_mels: 60 |
|
left_frames: 0 |
|
right_frames: 0 |
|
deltas: false |
|
|
|
|
|
out_n_neurons: 3 |
|
|
|
train_dataloader_options: |
|
num_workers: 0 |
|
batch_size: 32 |
|
|
|
val_dataloader_options: |
|
num_workers: 0 |
|
batch_size: 16 |
|
|
|
|
|
compute_features: &id003 !new:speechbrain.lobes.features.Fbank |
|
n_mels: 60 |
|
left_frames: 0 |
|
right_frames: 0 |
|
deltas: false |
|
|
|
embedding_model: &id004 !new:speechbrain.lobes.models.ECAPA_TDNN.ECAPA_TDNN |
|
input_size: 60 |
|
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: &id005 !new:speechbrain.lobes.models.Xvector.Classifier |
|
input_shape: [null, null, 256] |
|
activation: !name:torch.nn.LeakyReLU |
|
lin_blocks: 1 |
|
lin_neurons: 512 |
|
out_neurons: 3 |
|
|
|
epoch_counter: &id007 !new:speechbrain.utils.epoch_loop.EpochCounter |
|
limit: 40 |
|
|
|
|
|
augment_speed: &id001 !new:speechbrain.lobes.augment.TimeDomainSpecAugment |
|
sample_rate: 16000 |
|
speeds: [90, 100, 110] |
|
|
|
|
|
add_rev_noise: &id002 !new:speechbrain.lobes.augment.EnvCorrupt |
|
openrir_folder: ./ |
|
openrir_max_noise_len: 3.0 |
|
reverb_prob: 0.5 |
|
noise_prob: 0.8 |
|
noise_snr_low: 0 |
|
noise_snr_high: 15 |
|
rir_scale_factor: 1.0 |
|
|
|
|
|
|
|
|
|
|
|
augment_pipeline: [*id001, *id002] |
|
|
|
concat_augment: false |
|
|
|
mean_var_norm: &id006 !new:speechbrain.processing.features.InputNormalization |
|
|
|
norm_type: sentence |
|
std_norm: false |
|
|
|
modules: |
|
compute_features: *id003 |
|
augment_speed: *id001 |
|
add_rev_noise: *id002 |
|
embedding_model: *id004 |
|
classifier: *id005 |
|
mean_var_norm: *id006 |
|
compute_cost: !name:speechbrain.nnet.losses.nll_loss |
|
|
|
|
|
opt_class: !name:torch.optim.Adam |
|
lr: 0.001 |
|
weight_decay: 0.000002 |
|
|
|
lr_annealing: !new:speechbrain.nnet.schedulers.LinearScheduler |
|
initial_value: 0.001 |
|
final_value: 0.0001 |
|
epoch_count: 40 |
|
|
|
|
|
train_logger: !new:speechbrain.utils.train_logger.FileTrainLogger |
|
save_file: results_3lang/epaca/1988/train_log.txt |
|
|
|
|
|
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: results_3lang/epaca/1988/save |
|
recoverables: |
|
embedding_model: *id004 |
|
classifier: *id005 |
|
normalizer: *id006 |
|
counter: *id007 |
|
|