espnet/dongwei_ami_vad_rnn

Environments

date: Thu May 4 10:25:40 EDT 2023
python version: 3.8.16 (default, Mar 2 2023, 03:21:46) [GCC 11.2.0]
espnet version: espnet 202301
pytorch version: pytorch 1.8.1
Git hash: 1bd1db914b21bfb5ae5acbe2fc7162e3815ed260
- Commit date: Thu May 4 08:48:15 2023 -0400

Model info

Model link: https://huggingface.co/espnet/dongwei_ami_vad_rnn
ASR config: conf/tuning/train_vad_rnn.yaml
Decode config: conf/tuning/decode_rnn.yaml

exp/vad_train_asr_transformer_raw

PRECISION

dataset	value
exp/vad_train_asr_transformer_raw/decode_rnn_vad_model_valid.acc.ave/ihm_dev/result.txt	0.9311
exp/vad_train_asr_transformer_raw/decode_rnn_vad_model_valid.acc.ave/ihm_eval/result.txt	0.9547

RECALL

dataset	value
exp/vad_train_asr_transformer_raw/decode_rnn_vad_model_valid.acc.ave/ihm_dev/result.txt	0.9277
exp/vad_train_asr_transformer_raw/decode_rnn_vad_model_valid.acc.ave/ihm_eval/result.txt	0.9412

F1_SCORE

dataset	value
exp/vad_train_asr_transformer_raw/decode_rnn_vad_model_valid.acc.ave/ihm_dev/result.txt	0.9294
exp/vad_train_asr_transformer_raw/decode_rnn_vad_model_valid.acc.ave/ihm_eval/result.txt	0.9479

VAD config

expand

config: conf/tuning/train_vad_rnn.yaml
print_config: false
log_level: INFO
dry_run: false
iterator_type: sequence
output_dir: exp/vad_train_vad_rnn_raw
ngpu: 1
seed: 0
num_workers: 3
num_att_plot: 3
dist_backend: nccl
dist_init_method: env://
dist_world_size: null
dist_rank: null
local_rank: 0
dist_master_addr: null
dist_master_port: null
dist_launcher: null
multiprocessing_distributed: false
unused_parameters: false
sharded_ddp: false
cudnn_enabled: true
cudnn_benchmark: false
cudnn_deterministic: true
collect_stats: false
write_collected_feats: false
max_epoch: 2
patience: null
val_scheduler_criterion:
- valid
- loss
early_stopping_criterion:
- valid
- loss
- min
best_model_criterion:
-   - valid
    - acc
    - max
keep_nbest_models: 5
nbest_averaging_interval: 0
grad_clip: 5.0
grad_clip_type: 2.0
grad_noise: false
accum_grad: 1
no_forward_run: false
resume: true
train_dtype: float32
use_amp: false
log_interval: null
use_matplotlib: true
use_tensorboard: true
create_graph_in_tensorboard: false
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: 20
valid_batch_size: null
batch_bins: 14000000
valid_batch_bins: null
train_shape_file:
- exp/vad_stats_raw/train/speech_shape
- exp/vad_stats_raw/train/text_shape
valid_shape_file:
- exp/vad_stats_raw/valid/speech_shape
- exp/vad_stats_raw/valid/text_shape
batch_type: numel
valid_batch_type: null
fold_length:
- 80000
- 150
sort_in_batch: descending
sort_batch: descending
multiple_iterator: false
chunk_length: 500
chunk_shift_ratio: 0.5
num_cache_chunks: 1024
chunk_excluded_key_prefixes: []
train_data_path_and_name_and_type:
-   - dump/raw/ihm_train/wav.scp
    - speech
    - sound
-   - dump/raw/ihm_train/text
    - text
    - text
valid_data_path_and_name_and_type:
-   - dump/raw/ihm_dev/wav.scp
    - speech
    - sound
-   - dump/raw/ihm_dev/text
    - text
    - text
allow_variable_data_keys: false
max_cache_size: 0.0
max_cache_fd: 32
valid_max_cache_size: null
exclude_weight_decay: false
exclude_weight_decay_conf: {}
optim: adam
optim_conf:
    lr: 0.003
scheduler: warmuplr
scheduler_conf:
    warmup_steps: 25000
pre_postencoder_norm: false
init: null
input_size: null
use_preprocessor: true
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'
short_noise_thres: 0.5
segment_length: 10.0
frontend: default
frontend_conf:
    n_fft: 512
    win_length: 400
    hop_length: 160
    fs: 16k
specaug: specaug
specaug_conf:
    apply_time_warp: true
    time_warp_window: 5
    time_warp_mode: bicubic
    apply_freq_mask: true
    freq_mask_width_range:
    - 0
    - 30
    num_freq_mask: 2
    apply_time_mask: true
    time_mask_width_range:
    - 0
    - 40
    num_time_mask: 2
normalize: global_mvn
normalize_conf:
    stats_file: exp/vad_stats_raw/train/feats_stats.npz
model: espnet
model_conf:
    length_normalized_loss: false
preencoder: null
preencoder_conf: {}
encoder: rnn
encoder_conf:
    rnn_type: gru
    bidirectional: true
    use_projection: true
    num_layers: 4
    hidden_size: 320
    output_size: 320
    dropout: 0.2
    subsample:
    - 1
    - 1
    - 1
    - 1
required:
- output_dir
version: '202304'
distributed: false

Citing ESPnet

@inproceedings{watanabe2018espnet,
  author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
  title={{ESPnet}: End-to-End Speech Processing Toolkit},
  year={2018},
  booktitle={Proceedings of Interspeech},
  pages={2207--2211},
  doi={10.21437/Interspeech.2018-1456},
  url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
}

or arXiv:

@misc{watanabe2018espnet,
  title={ESPnet: End-to-End Speech Processing Toolkit},
  author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
  year={2018},
  eprint={1804.00015},
  archivePrefix={arXiv},
  primaryClass={cs.CL}
}