|
--- |
|
tags: |
|
- espnet |
|
- audio |
|
- automatic-speech-recognition |
|
language: noinfo |
|
datasets: |
|
- yolo_mixtec |
|
license: cc-by-4.0 |
|
--- |
|
|
|
## ESPnet2 ASR model |
|
|
|
### `espnet/ftshijt_espnet2_asr_yolo_mixtec_transformer` |
|
|
|
This model was trained by ftshijt using yolo_mixtec recipe in [espnet](https://github.com/espnet/espnet/). |
|
|
|
### Demo: How to use in ESPnet2 |
|
|
|
```bash |
|
cd espnet |
|
|
|
pip install -e . |
|
cd els/yolo_mixtec/asr1 |
|
./run.sh --skip_data_prep false --skip_train true --download_model espnet/ftshijt_espnet2_asr_yolo_mixtec_transformer |
|
``` |
|
|
|
<!-- Generated by scripts/utils/show_asr_result.sh --> |
|
# RESULTS |
|
## Environments |
|
- date: `Wed Nov 10 02:59:39 EST 2021` |
|
- python version: `3.9.7 (default, Sep 16 2021, 13:09:58) [GCC 7.5.0]` |
|
- espnet version: `espnet 0.10.4a1` |
|
- pytorch version: `pytorch 1.9.0` |
|
- Git hash: `` |
|
- Commit date: `` |
|
|
|
## asr_train_asr_transformer_specaug_raw_bpe500 |
|
### WER |
|
|
|
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |
|
|---|---|---|---|---|---|---|---|---| |
|
|decode_asr_lm_lm_train_bpe500_valid.loss.ave_asr_model_valid.acc.best/test|4985|81348|84.1|11.8|4.1|2.5|18.3|82.5| |
|
|
|
### CER |
|
|
|
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |
|
|---|---|---|---|---|---|---|---|---| |
|
|decode_asr_lm_lm_train_bpe500_valid.loss.ave_asr_model_valid.acc.best/test|4985|626187|93.4|2.2|4.4|2.4|9.0|82.5| |
|
|
|
### TER |
|
|
|
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |
|
|---|---|---|---|---|---|---|---|---| |
|
|decode_asr_lm_lm_train_bpe500_valid.loss.ave_asr_model_valid.acc.best/test|4985|325684|90.7|5.2|4.1|2.2|11.5|82.5| |
|
|
|
## ASR config |
|
|
|
<details><summary>expand</summary> |
|
|
|
``` |
|
config: conf/tuning/train_asr_transformer_specaug.yaml |
|
print_config: false |
|
log_level: INFO |
|
dry_run: false |
|
iterator_type: sequence |
|
output_dir: exp/asr_train_asr_transformer_specaug_raw_bpe500 |
|
ngpu: 1 |
|
seed: 0 |
|
num_workers: 1 |
|
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: 100 |
|
patience: 15 |
|
val_scheduler_criterion: |
|
- valid |
|
- loss |
|
early_stopping_criterion: |
|
- valid |
|
- loss |
|
- min |
|
best_model_criterion: |
|
- - valid |
|
- acc |
|
- max |
|
keep_nbest_models: 10 |
|
grad_clip: 5 |
|
grad_clip_type: 2.0 |
|
grad_noise: false |
|
accum_grad: 2 |
|
no_forward_run: false |
|
resume: true |
|
train_dtype: float32 |
|
use_amp: false |
|
log_interval: null |
|
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: 32 |
|
valid_batch_size: null |
|
batch_bins: 1000000 |
|
valid_batch_bins: null |
|
train_shape_file: |
|
- exp/asr_stats_raw_bpe500/train/speech_shape |
|
- exp/asr_stats_raw_bpe500/train/text_shape.bpe |
|
valid_shape_file: |
|
- exp/asr_stats_raw_bpe500/valid/speech_shape |
|
- exp/asr_stats_raw_bpe500/valid/text_shape.bpe |
|
batch_type: folded |
|
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 |
|
train_data_path_and_name_and_type: |
|
- - /tmp/st-jiatong-54826.tbQP9L0N/raw/train/wav.scp |
|
- speech |
|
- kaldi_ark |
|
- - /tmp/st-jiatong-54826.tbQP9L0N/raw/train/text |
|
- text |
|
- text |
|
valid_data_path_and_name_and_type: |
|
- - /tmp/st-jiatong-54826.tbQP9L0N/raw/dev/wav.scp |
|
- speech |
|
- kaldi_ark |
|
- - /tmp/st-jiatong-54826.tbQP9L0N/raw/dev/text |
|
- text |
|
- text |
|
allow_variable_data_keys: false |
|
max_cache_size: 0.0 |
|
max_cache_fd: 32 |
|
valid_max_cache_size: null |
|
optim: adam |
|
optim_conf: |
|
lr: 1.0 |
|
scheduler: noamlr |
|
scheduler_conf: |
|
warmup_steps: 25000 |
|
token_list: |
|
- <blank> |
|
- <unk> |
|
- '4' |
|
- '3' |
|
- '1' |
|
- '2' |
|
- A |
|
- ▁NDI |
|
- '''4' |
|
- '''1' |
|
- U |
|
- ▁BA |
|
- O |
|
- ▁I |
|
- E |
|
- 4= |
|
- ▁KU |
|
- ▁TAN |
|
- ▁KA |
|
- '''3' |
|
- NI |
|
- ▁YA |
|
- RA |
|
- 3= |
|
- 2= |
|
- IN |
|
- NA |
|
- ▁TA |
|
- AN |
|
- ▁KAN |
|
- ▁NI |
|
- ▁NDA |
|
- ▁NA |
|
- ▁JI |
|
- KAN |
|
- CHI |
|
- (3)= |
|
- I |
|
- UN |
|
- 1- |
|
- ▁SA |
|
- (4)= |
|
- ▁JA |
|
- XI |
|
- ▁KO |
|
- ▁TI |
|
- TA |
|
- KU |
|
- BI |
|
- ▁YU |
|
- ▁KWA |
|
- KA |
|
- XA |
|
- 1= |
|
- ▁YO |
|
- RI |
|
- NDO |
|
- ▁XA |
|
- TU |
|
- ▁TU |
|
- ▁ÑA |
|
- ▁KI |
|
- ▁XI |
|
- YO |
|
- NDU |
|
- NDA |
|
- ▁CHI |
|
- (2)= |
|
- ▁BI |
|
- ▁NU |
|
- KI |
|
- (1)= |
|
- YU |
|
- 3- |
|
- ▁MI |
|
- 'ON' |
|
- ▁A |
|
- BA |
|
- 4- |
|
- KO |
|
- ▁NDU |
|
- ▁ÑU |
|
- ▁NDO |
|
- NU |
|
- ÑU |
|
- '143' |
|
- ▁SI |
|
- ▁SO |
|
- 13- |
|
- NDI |
|
- ▁AN |
|
- ▁SU |
|
- TIN |
|
- SA |
|
- ▁BE |
|
- TO |
|
- RUN |
|
- KWA |
|
- KWI |
|
- ▁NDE |
|
- ▁KWI |
|
- XIN |
|
- ▁U |
|
- SI |
|
- SO |
|
- ▁TUN |
|
- EN |
|
- ▁KWE |
|
- YA |
|
- (4)=2 |
|
- NDE |
|
- TI |
|
- TUN |
|
- ▁TIN |
|
- MA |
|
- ▁SE |
|
- ▁XU |
|
- SU |
|
- ▁LU |
|
- ▁KE |
|
- ▁ |
|
- MI |
|
- ▁RAN |
|
- (3)=2 |
|
- 14- |
|
- ▁MA |
|
- KUN |
|
- LU |
|
- N |
|
- ▁O |
|
- KE |
|
- NGA |
|
- ▁IS |
|
- ▁JU |
|
- '=' |
|
- ▁LA |
|
- ÑA |
|
- JA |
|
- CHUN |
|
- R |
|
- TAN |
|
- PU |
|
- ▁TIEM |
|
- LI |
|
- LA |
|
- CHIU |
|
- ▁PA |
|
- M |
|
- ▁REY |
|
- ▁BAN |
|
- JI |
|
- L |
|
- SUN |
|
- ▁SEÑOR |
|
- ▁JO |
|
- ▁TIO |
|
- KWE |
|
- CHU |
|
- S |
|
- ▁YE |
|
- KIN |
|
- XU |
|
- BE |
|
- ▁CUENTA |
|
- ▁SAN |
|
- RRU |
|
- ▁¿ |
|
- CHA |
|
- ▁TO |
|
- RRA |
|
- LO |
|
- TE |
|
- ▁AMIGU |
|
- PA |
|
- XAN |
|
- ▁C |
|
- C |
|
- ▁CHA |
|
- ▁TE |
|
- ▁HIJO |
|
- ▁MB |
|
- ▁PI |
|
- G |
|
- ▁ÁNIMA |
|
- ▁CHE |
|
- ▁P |
|
- B |
|
- NDIO |
|
- SE |
|
- ▁SANTU |
|
- MU |
|
- ▁PADRE |
|
- D |
|
- JU |
|
- Z |
|
- ▁TORO |
|
- ▁PO |
|
- LE |
|
- ▁LI |
|
- RO |
|
- ▁LO |
|
- ▁MESA |
|
- CA |
|
- ▁CHIU |
|
- DO |
|
- ▁BU |
|
- ▁BUTA |
|
- JO |
|
- T |
|
- TRU |
|
- RU |
|
- ▁MBO |
|
- ▁JUAN |
|
- ▁MM |
|
- ▁CA |
|
- ▁M |
|
- ▁MAS |
|
- ▁DE |
|
- V |
|
- ▁MAÑA |
|
- ▁UTA |
|
- DA |
|
- ▁MULA |
|
- ▁YOLOXÓCHITL |
|
- ▁CONSEJU |
|
- ▁Y |
|
- ▁LE |
|
- ÓN |
|
- ▁MISA |
|
- TIU |
|
- ▁CANDELA |
|
- ▁PATRÓN |
|
- ▁PADRINU |
|
- ▁MARCU |
|
- ▁V |
|
- ▁G |
|
- Í |
|
- ▁XE |
|
- ▁MU |
|
- ▁XO |
|
- NGUI |
|
- ▁CO |
|
- ▁HOMBRE |
|
- ▁PESU |
|
- ▁PE |
|
- ▁D |
|
- ▁MACHITI |
|
- CO |
|
- REN |
|
- ▁RANCHU |
|
- ▁MIS |
|
- ▁MACHU |
|
- J |
|
- ▁PAN |
|
- CHO |
|
- H |
|
- ▁CHU |
|
- Y |
|
- ▁TON |
|
- GA |
|
- X |
|
- ▁VI |
|
- ▁FE |
|
- ▁TARRAYA |
|
- ▁SANTÍSIMA |
|
- ▁N |
|
- ▁MAYÓ |
|
- ▁CARRU |
|
- ▁F |
|
- ▁PAPÁ |
|
- ▁PALOMA |
|
- ▁MARÍA |
|
- ▁PEDRU |
|
- ▁CAFÉ |
|
- ▁COMISARIO |
|
- ▁PANELA |
|
- ▁PELÓN |
|
- É |
|
- ▁POZO |
|
- ▁CABRÓN |
|
- ▁GUACHU |
|
- ▁S |
|
- RES |
|
- ▁COSTUMBRE |
|
- ▁SEÑA |
|
- QUI |
|
- ▁ORO |
|
- CH |
|
- ▁MAR |
|
- SIN |
|
- SAN |
|
- ▁COSTA |
|
- ▁MAMÁ |
|
- ▁CINCUENTA |
|
- ▁CHO |
|
- ▁PEDR |
|
- ▁JUNTA |
|
- MÚ |
|
- ▁TIENDA |
|
- ▁JOSÉ |
|
- NC |
|
- ▁ES |
|
- ▁SUERTE |
|
- ▁FAMILIA |
|
- ▁ZAPATU |
|
- NTE |
|
- ▁PASTO |
|
- ▁CON |
|
- Ñ |
|
- ▁BOTE |
|
- CIÓN |
|
- ▁RE |
|
- ▁BOLSA |
|
- ▁MANGO |
|
- ▁JWE |
|
- ▁GASTU |
|
- ▁T |
|
- ▁B |
|
- ▁KW |
|
- ÍN |
|
- ▁HIJA |
|
- ▁CUARENT |
|
- ▁VAQUERU |
|
- ▁NECHITO |
|
- ▁NOVIA |
|
- ▁NOVIO |
|
- JWE |
|
- ▁PUENTE |
|
- ▁SANDÍA |
|
- ▁MALA |
|
- Ó |
|
- ▁ABONO |
|
- ▁JESÚS |
|
- ▁CUARTO |
|
- ▁EFE |
|
- ▁REINA |
|
- ▁COMANDANTE |
|
- ▁ESCUELA |
|
- ▁MANZANA |
|
- ▁MÁQUINA |
|
- LLA |
|
- ▁COR |
|
- ▁JERÓNIMO |
|
- ▁PISTOLA |
|
- NGI |
|
- CIO |
|
- ▁FRANCISCU |
|
- ▁TEODORO |
|
- CER |
|
- ▁SALUBI |
|
- ▁MEZA |
|
- ▁MÚSIC |
|
- ▁RU |
|
- ▁CONSTANTINO |
|
- ▁GARCÍA |
|
- ▁FRENU |
|
- ▁ROSA |
|
- ▁CERVEZA |
|
- ▁CIGARRU |
|
- ▁COMISIÓN |
|
- ▁CUNIJO |
|
- ▁FRANCISCO |
|
- ▁HÍJOLE |
|
- ▁NUEVE |
|
- ▁MUL |
|
- ▁PANTALÓN |
|
- ▁CAMISA |
|
- ▁CHINGADA |
|
- ▁SEMANA |
|
- ▁COM |
|
- GAR |
|
- ▁MARTÍN |
|
- ▁SÁBADO |
|
- ▁TRABAJO |
|
- ▁CINCO |
|
- ▁DIE |
|
- ▁EST |
|
- NDWA |
|
- ▁LECHIN |
|
- ▁COCO |
|
- ILLU |
|
- ▁CORRE |
|
- ▁MADR |
|
- ▁REC |
|
- ▁BAUTISTA |
|
- ▁VENTANA |
|
- ▁CUÑAD |
|
- ▁ANTONIU |
|
- ▁COPALA |
|
- LÍN |
|
- ▁SECUND |
|
- ▁COHETE |
|
- ▁HISTORIA |
|
- ▁POLICÍA |
|
- ENCIA |
|
- ▁CAD |
|
- ▁LUIS |
|
- ▁DOCTOR |
|
- ▁GONZÁLEZ |
|
- ▁JUEVE |
|
- ▁LIBRU |
|
- ▁QUESU |
|
- ▁VIAJE |
|
- ▁CART |
|
- ▁LOCO |
|
- ▁BOL |
|
- ▁COMPADRE |
|
- ▁JWI |
|
- ▁METRU |
|
- ▁BUENO |
|
- ▁TRE |
|
- ▁CASTILLO |
|
- ▁COMITÉ |
|
- ▁ETERNO |
|
- ▁LÍQUIDO |
|
- ▁MOLE |
|
- ▁CAPULCU |
|
- ▁DOMING |
|
- ▁ROMA |
|
- ▁CARAJU |
|
- ▁RIATA |
|
- ▁TRATU |
|
- ▁SEIS |
|
- ▁ADÁN |
|
- ▁JUANCITO |
|
- ▁HOR |
|
- '''' |
|
- ▁ARRÓ |
|
- ▁COCINA |
|
- ▁PALACIO |
|
- ▁RÓMULO |
|
- K |
|
- ▁ALFONSO |
|
- ▁BARTOLO |
|
- ▁FELIPE |
|
- ▁HERRER |
|
- ▁PAULINO |
|
- ▁YEGUA |
|
- ▁LISTA |
|
- Ú |
|
- ▁ABRIL |
|
- ▁CUATRO |
|
- ▁DICIEMBRE |
|
- ▁MARGARITO |
|
- ▁MOJONERA |
|
- ▁SOLEDAD |
|
- ▁VESTIDO |
|
- ▁PELOTA |
|
- RRET |
|
- ▁CAPITÁN |
|
- ▁COMUNIÓN |
|
- ▁CUCHARA |
|
- ▁FERNANDO |
|
- ▁GUADALUPE |
|
- ▁MIGUEL |
|
- ▁PELÚN |
|
- ▁SECRETARIU |
|
- ▁LENCHU |
|
- ▁EVA |
|
- ▁SEGUND |
|
- ▁CANTOR |
|
- ▁CHILPANCINGO |
|
- ▁GABRIEL |
|
- ▁QUINIENTO |
|
- ▁RAÚL |
|
- ▁SEVERIAN |
|
- ▁TUMBADA |
|
- ▁MALINCHI |
|
- ▁PRIMU |
|
- ▁MORAL |
|
- ▁AGOSTO |
|
- ▁CENTÍMETRO |
|
- ▁FIRMA |
|
- ▁HUEHUETÁN |
|
- ▁MANGUERA |
|
- ▁MEDI |
|
- ▁MUERT |
|
- ▁SALAZAR |
|
- ▁VIERNI |
|
- LILL |
|
- ▁LL |
|
- '-' |
|
- ▁CAMPESINO |
|
- ▁CIVIL |
|
- ▁COMISARIADO |
|
- ) |
|
- ( |
|
- Ã |
|
- ‘ |
|
- ¿ |
|
- Ü |
|
- ¡ |
|
- Q |
|
- F |
|
- Á |
|
- P |
|
- Ÿ |
|
- W |
|
- Ý |
|
- <sos/eos> |
|
init: xavier_uniform |
|
input_size: null |
|
ctc_conf: |
|
dropout_rate: 0.0 |
|
ctc_type: builtin |
|
reduce: true |
|
ignore_nan_grad: true |
|
model_conf: |
|
ctc_weight: 0.3 |
|
lsm_weight: 0.1 |
|
length_normalized_loss: false |
|
use_preprocessor: true |
|
token_type: bpe |
|
bpemodel: data/token_list/bpe_unigram500/bpe.model |
|
non_linguistic_symbols: null |
|
cleaner: null |
|
g2p: 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 |
|
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/asr_stats_raw_bpe500/train/feats_stats.npz |
|
preencoder: null |
|
preencoder_conf: {} |
|
encoder: transformer |
|
encoder_conf: |
|
input_layer: conv2d |
|
num_blocks: 12 |
|
linear_units: 2048 |
|
dropout_rate: 0.1 |
|
output_size: 512 |
|
attention_heads: 4 |
|
attention_dropout_rate: 0.0 |
|
postencoder: null |
|
postencoder_conf: {} |
|
decoder: transformer |
|
decoder_conf: |
|
input_layer: embed |
|
num_blocks: 6 |
|
linear_units: 2048 |
|
dropout_rate: 0.1 |
|
required: |
|
- output_dir |
|
- token_list |
|
version: 0.10.4a1 |
|
distributed: false |
|
``` |
|
|
|
</details> |
|
|
|
|
|
|
|
### Citing ESPnet |
|
|
|
```BibTex |
|
@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: |
|
|
|
```bibtex |
|
@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} |
|
} |
|
``` |
|
|