train.yaml

# ################################
# Model: wav2vec2 + DNN + CTC
# Augmentation: SpecAugment
# Authors: Titouan Parcollet 2021
# ################################

# Seed needs to be set at top of yaml, before objects with parameters are made
seed: 1234
__set_seed: !!python/object/apply:torch.manual_seed [!ref <seed>]
output_folder: !ref model/<seed>
wer_file: !ref <output_folder>/wer.txt
save_folder: !ref <output_folder>/save
train_log: !ref <output_folder>/train_log.txt

# URL for the biggest LeBenchmark wav2vec french.
wav2vec2_folder: !ref <save_folder>/wav2vec2_checkpoint

# Data files
data_folder: /path/to/data  # e.g, /localscratch/cv-corpus-5.1-2020-06-22/fr
train_tsv_file: !ref <data_folder>/train.tsv  # Standard CommonVoice .tsv files
dev_tsv_file: !ref <data_folder>/dev.tsv  # Standard CommonVoice .tsv files
test_tsv_file: !ref <data_folder>/test.tsv  # Standard CommonVoice .tsv files
accented_letters: True
language: fr # use 'it' for Italian, 'rw' for Kinyarwanda, 'en' for english
train_csv: Data/train_wavs/train.csv
valid_csv: Data/dev_wavs/dev.csv
test_csv: 
    - Data/test_wavs/test.csv

skip_prep: True # Skip data preparation

use_language_modelling: True
ngram_lm_path: languageModel.arpa

# We remove utterance slonger than 10s in the train/dev/test sets as
# longer sentences certainly correspond to "open microphones".
avoid_if_longer_than: 10.0
avoid_if_shorter_than: 1.2


# Training parameters
number_of_epochs: 12
lr: 1.0
lr_wav2vec: 0.0001
sorting: ascending
auto_mix_prec: False
sample_rate: 16000
ckpt_interval_minutes: 30 # save checkpoint every N min

# With data_parallel batch_size is split into N jobs
# With DDP batch_size is multiplied by N jobs
# Must be 6 per GPU to fit 16GB of VRAM
batch_size: 10
test_batch_size: 4

dataloader_options:
    batch_size: !ref <batch_size>
    num_workers: 6
test_dataloader_options:
    batch_size: !ref <test_batch_size>
    num_workers: 6

# BPE parameters
token_type: char  # ["unigram", "bpe", "char"]
character_coverage: 1.0

# Model parameters
# activation: !name:torch.nn.LeakyReLU
wav2vec_output_dim: 1024
dnn_neurons: 1024
freeze_wav2vec: False
freeze_feature_extractor: True
dropout: 0.15
warmup_steps: 500 # The wav2vec 2 model isn't updated for this amount of steps

# Outputs
output_neurons: 40  # BPE size, index(blank/eos/bos) = 0

# Decoding parameters
# Be sure that the bos and eos index match with the BPEs ones
blank_index: 0
unk_index: 1

#
# Functions and classes
#
epoch_counter: !new:speechbrain.utils.epoch_loop.EpochCounter
    limit: !ref <number_of_epochs>


enc: !new:speechbrain.nnet.containers.Sequential
    input_shape: [null, null, !ref <wav2vec_output_dim>]
    linear1: !name:speechbrain.nnet.linear.Linear
        n_neurons: !ref <dnn_neurons>
        bias: True
    bn1: !name:speechbrain.nnet.normalization.BatchNorm1d
    activation: !new:torch.nn.LeakyReLU
    drop: !new:torch.nn.Dropout
        p: !ref <dropout>
    linear2: !name:speechbrain.nnet.linear.Linear
        n_neurons: !ref <dnn_neurons>
        bias: True
    bn2: !name:speechbrain.nnet.normalization.BatchNorm1d
    activation2: !new:torch.nn.LeakyReLU
    drop2: !new:torch.nn.Dropout
        p: !ref <dropout>
    linear3: !name:speechbrain.nnet.linear.Linear
        n_neurons: !ref <dnn_neurons>
        bias: True
    bn3: !name:speechbrain.nnet.normalization.BatchNorm1d
    activation3: !new:torch.nn.LeakyReLU

wav2vec2: !new:speechbrain.lobes.models.huggingface_transformers.wav2vec2.Wav2Vec2
    source: wavlm-large/
    output_norm: False
    freeze: !ref <freeze_wav2vec>
    freeze_feature_extractor: !ref <freeze_feature_extractor>
    save_path: !ref <wav2vec2_folder>


ctc_lin: !new:speechbrain.nnet.linear.Linear
    input_size: !ref <dnn_neurons>
    n_neurons: !ref <output_neurons>

log_softmax: !new:speechbrain.nnet.activations.Softmax
    apply_log: True

ctc_cost: !name:speechbrain.nnet.losses.ctc_loss
    blank_index: !ref <blank_index>

modules:
    wav2vec2: !ref <wav2vec2>
    enc: !ref <enc>
    ctc_lin: !ref <ctc_lin>

model: !new:torch.nn.ModuleList
    - [!ref <enc>, !ref <ctc_lin>]

model_opt_class: !name:torch.optim.Adadelta
    lr: !ref <lr>
    rho: 0.95
    eps: 1.e-8

wav2vec_opt_class: !name:torch.optim.Adam
    lr: !ref <lr_wav2vec>

lr_annealing_model: !new:speechbrain.nnet.schedulers.NewBobScheduler
    initial_value: !ref <lr>
    improvement_threshold: 0.0025
    annealing_factor: 0.8
    patient: 0

lr_annealing_wav2vec: !new:speechbrain.nnet.schedulers.NewBobScheduler
    initial_value: !ref <lr_wav2vec>
    improvement_threshold: 0.0025
    annealing_factor: 0.9
    patient: 0

checkpointer: !new:speechbrain.utils.checkpoints.Checkpointer
    checkpoints_dir: !ref <save_folder>
    recoverables:
        wav2vec2: !ref <wav2vec2>
        model: !ref <model>
        scheduler_model: !ref <lr_annealing_model>
        scheduler_wav2vec: !ref <lr_annealing_wav2vec>
        counter: !ref <epoch_counter>

train_logger: !new:speechbrain.utils.train_logger.FileTrainLogger
    save_file: !ref <train_log>

error_rate_computer: !name:speechbrain.utils.metric_stats.ErrorRateStats

cer_computer: !name:speechbrain.utils.metric_stats.ErrorRateStats
    split_tokens: True