exp/bestrq_train_bestrq_conformerL_maskprob1e-2_unmaskedonly_l2norm_cb8192_accum8_bs256_amp_lr5e-4_seed20220404_chunk3k_raw_en/config.yaml

config: conf/tuning/train_bestrq_conformerL_maskprob1e-2_unmaskedonly_l2norm_cb8192_accum8_bs256_amp_lr5e-4_seed20220404_chunk3k.yaml
print_config: false
log_level: INFO
dry_run: false
iterator_type: sequence
output_dir: exp/bestrq_train_bestrq_conformerL_maskprob1e-2_unmaskedonly_l2norm_cb8192_accum8_bs256_amp_lr5e-4_seed20220404_chunk3k_raw_en
ngpu: 1
seed: 0
num_workers: 1
num_att_plot: 3
dist_backend: nccl
dist_init_method: env://
dist_world_size: 4
dist_rank: 0
local_rank: 0
dist_master_addr: localhost
dist_master_port: 54869
dist_launcher: null
multiprocessing_distributed: true
unused_parameters: false
sharded_ddp: false
cudnn_enabled: true
cudnn_benchmark: false
cudnn_deterministic: false
collect_stats: false
write_collected_feats: false
max_epoch: 500
patience: null
val_scheduler_criterion:
- valid
- loss
early_stopping_criterion:
- valid
- loss
- min
best_model_criterion:
-   - valid
    - masked_acc
    - max
keep_nbest_models: 300
nbest_averaging_interval: 0
grad_clip: 5.0
grad_clip_type: 2.0
grad_noise: false
accum_grad: 8
no_forward_run: false
resume: true
train_dtype: float32
use_amp: true
log_interval: null
use_matplotlib: true
use_tensorboard: true
use_wandb: false
wandb_project: null
wandb_id: null
wandb_entity: null
wandb_name: null
wandb_model_log_interval: -1
detect_anomaly: false
pretrain_path: null
init_param: []
ignore_init_mismatch: false
freeze_param: []
num_iters_per_epoch: null
batch_size: 256
valid_batch_size: null
batch_bins: 1000000
valid_batch_bins: null
train_shape_file:
- exp/bestrq_stats_raw_en/train/speech_shape
valid_shape_file:
- exp/bestrq_stats_raw_en/valid/speech_shape
batch_type: folded
valid_batch_type: null
fold_length:
- 80000
sort_in_batch: descending
sort_batch: descending
multiple_iterator: false
chunk_length: 500
chunk_shift_ratio: 0.5
num_cache_chunks: 1024
train_data_path_and_name_and_type:
-   - dump/raw/train_6k/wav.scp
    - speech
    - sound
valid_data_path_and_name_and_type:
-   - dump/raw/dev/wav.scp
    - speech
    - sound
allow_variable_data_keys: false
max_cache_size: 0.0
max_cache_fd: 32
valid_max_cache_size: null
optim: adam
optim_conf:
    lr: 0.0005
scheduler: warmuplr
scheduler_conf:
    warmup_steps: 30000
init: xavier_uniform
input_size: null
model_conf:
    lsm_weight: 0.0
    length_normalized_loss: false
    vector_size: 16
    codebook_size: 8192
    temporal_reduction: 4
    mask_prob: 0.01
    mask_length: 40
    unmasked_region_weight: 0
    apply_l2_normalization: true
    codebook_and_matrix_init_file: data/local/codebook_and_matrix_seed20220404.pth
    frame_chunk_size: 3200
use_preprocessor: true
token_type: null
cleaner: null
speech_volume_normalize: null
rir_scp: null
rir_apply_prob: 1.0
noise_scp: null
noise_apply_prob: 1.0
noise_db_range: '13_15'
frontend: default
frontend_conf:
    fs: 16k
    n_fft: 512
    win_length: 400
    hop_length: 160
    frontend_conf: null
    apply_stft: true
specaug: null
specaug_conf: {}
normalize: global_mvn
normalize_conf:
    stats_file: exp/bestrq_stats_raw_en/train/feats_stats.npz
preencoder: null
preencoder_conf: {}
encoder: conformer
encoder_conf:
    output_size: 512
    attention_heads: 8
    linear_units: 2048
    num_blocks: 17
    dropout_rate: 0.1
    positional_dropout_rate: 0.1
    attention_dropout_rate: 0.1
    input_layer: conv2d
    normalize_before: true
    macaron_style: true
    rel_pos_type: latest
    pos_enc_layer_type: rel_pos
    selfattention_layer_type: rel_selfattn
    activation_type: swish
    use_cnn_module: true
    cnn_module_kernel: 31
postencoder: null
postencoder_conf: {}
required:
- output_dir
version: 0.10.7a1
distributed: true