xlsr-53-korean_kids_3_to_5_syl / syl_nospace_50epoch_2nd.yaml
lsnoo's picture
YAML file used for training the model
46a8e81
raw
history blame
1.2 kB
# @package _group_
common:
fp16: true
log_format: tqdm
log_interval: 200
tensorboard_logdir: /workspace/data/finetuning/syl_nospace_50epoch_2nd/logs/
checkpoint:
no_epoch_checkpoints: true
best_checkpoint_metric: wer
save_dir: /workspace/data/finetuning/syl_nospace_50epoch_2nd/checkpoints/
task:
_name: audio_finetuning
data: /workspace/data/finetuning/syl_nospace_50epoch_2nd
normalize: true
labels: ltr
dataset:
num_workers: 4
max_tokens: 500000
skip_invalid_size_inputs_valid_test: true
valid_subset: valid
distributed_training:
ddp_backend: legacy_ddp
distributed_world_size: 1
criterion:
_name: ctc
zero_infinity: true
optimization:
max_epoch: 50
max_update: 800000
lr: [0.00003]
sentence_avg: true
update_freq: [4]
optimizer:
_name: adam
adam_betas: (0.9,0.98)
adam_eps: 1e-08
lr_scheduler:
_name: tri_stage
phase_ratio: [0.1, 0.4, 0.5]
final_lr_scale: 0.05
model:
_name: wav2vec_ctc
w2v_path: /workspace/data/finetuning/xlsr_53_56k.pt
apply_mask: true
mask_prob: 0.65
mask_channel_prob: 0.5
mask_channel_length: 64
layerdrop: 0.1
activation_dropout: 0.1
feature_grad_mult: 0.0
freeze_finetune_updates: 0