|
--- |
|
tags: |
|
- espnet |
|
- audio |
|
- automatic-speech-recognition |
|
language: es |
|
datasets: |
|
- fisher_callhome_spanish |
|
license: cc-by-4.0 |
|
--- |
|
|
|
## ESPnet2 ASR model |
|
|
|
### `pyf98/fisher_callhome_spanish_e_branchformer` |
|
|
|
This model was trained by Yifan Peng using fisher_callhome_spanish recipe in [espnet](https://github.com/espnet/espnet/). |
|
|
|
References: |
|
- [E-Branchformer: Branchformer with Enhanced merging for speech recognition (SLT 2022)](https://arxiv.org/abs/2210.00077) |
|
- [Branchformer: Parallel MLP-Attention Architectures to Capture Local and Global Context for Speech Recognition and Understanding (ICML 2022)](https://proceedings.mlr.press/v162/peng22a.html) |
|
|
|
### Demo: How to use in ESPnet2 |
|
|
|
Follow the [ESPnet installation instructions](https://espnet.github.io/espnet/installation.html) |
|
if you haven't done that already. |
|
|
|
```bash |
|
cd espnet |
|
git checkout 568bd0808f7509f9735282537db4c68dc3bdf376 |
|
pip install -e . |
|
cd egs2/fisher_callhome_spanish/asr1 |
|
./run.sh --skip_data_prep false --skip_train true --download_model pyf98/fisher_callhome_spanish_e_branchformer |
|
``` |
|
|
|
<!-- Generated by scripts/utils/show_asr_result.sh --> |
|
# RESULTS |
|
## Environments |
|
- date: `Tue Feb 28 21:03:54 CST 2023` |
|
- python version: `3.9.15 (main, Nov 24 2022, 14:31:59) [GCC 11.2.0]` |
|
- espnet version: `espnet 202301` |
|
- pytorch version: `pytorch 1.13.1` |
|
- Git hash: `568bd0808f7509f9735282537db4c68dc3bdf376` |
|
- Commit date: `Tue Feb 28 06:06:06 2023 -0500` |
|
|
|
## exp/asr_train_asr_e_branchformer_e16_raw_bpe1000_sp |
|
### WER |
|
|
|
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |
|
|---|---|---|---|---|---|---|---|---| |
|
|decode_conformer_asr_model_valid.acc.ave/callhome_devtest|3964|37989|69.4|23.2|7.5|7.2|37.8|79.0| |
|
|decode_conformer_asr_model_valid.acc.ave/callhome_evltest|1829|19035|68.6|23.1|8.2|6.3|37.6|81.7| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_dev|3979|40961|83.7|11.9|4.5|4.2|20.5|62.4| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_dev2|3961|39888|84.4|11.8|3.9|4.6|20.2|62.8| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_test|3641|40011|86.3|10.2|3.5|5.0|18.7|60.3| |
|
|
|
### CER |
|
|
|
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |
|
|---|---|---|---|---|---|---|---|---| |
|
|decode_conformer_asr_model_valid.acc.ave/callhome_devtest|3964|181052|84.4|6.4|9.3|6.4|22.0|79.0| |
|
|decode_conformer_asr_model_valid.acc.ave/callhome_evltest|1829|91266|83.7|6.4|9.8|5.7|21.9|81.7| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_dev|3979|194297|93.2|2.6|4.2|3.9|10.7|62.4| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_dev2|3961|189965|93.8|2.5|3.6|4.1|10.3|62.8| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_test|3641|194507|94.8|2.1|3.1|4.6|9.8|60.3| |
|
|
|
### TER |
|
|
|
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |
|
|---|---|---|---|---|---|---|---|---| |
|
|decode_conformer_asr_model_valid.acc.ave/callhome_devtest|3964|57692|66.2|18.6|15.2|5.0|38.8|79.0| |
|
|decode_conformer_asr_model_valid.acc.ave/callhome_evltest|1829|28951|65.5|18.2|16.3|4.8|39.3|81.7| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_dev|3979|55907|83.4|9.7|6.9|3.9|20.5|62.4| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_dev2|3961|53966|84.5|9.6|5.8|4.2|19.6|62.8| |
|
|decode_conformer_asr_model_valid.acc.ave/fisher_test|3641|54212|86.8|8.2|5.0|4.8|18.0|60.3| |
|
|
|
## ASR config |
|
|
|
<details><summary>expand</summary> |
|
|
|
``` |
|
config: conf/tuning/train_asr_e_branchformer_e16.yaml |
|
print_config: false |
|
log_level: INFO |
|
dry_run: false |
|
iterator_type: sequence |
|
output_dir: exp/asr_train_asr_e_branchformer_e16_raw_bpe1000_sp |
|
ngpu: 1 |
|
seed: 0 |
|
num_workers: 4 |
|
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: 50 |
|
patience: null |
|
val_scheduler_criterion: |
|
- valid |
|
- loss |
|
early_stopping_criterion: |
|
- valid |
|
- loss |
|
- min |
|
best_model_criterion: |
|
- - valid |
|
- acc |
|
- max |
|
keep_nbest_models: 10 |
|
nbest_averaging_interval: 0 |
|
grad_clip: 3 |
|
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: 10000000 |
|
valid_batch_bins: null |
|
train_shape_file: |
|
- exp/asr_stats_raw_bpe1000_sp/train/speech_shape |
|
- exp/asr_stats_raw_bpe1000_sp/train/text_shape.bpe |
|
valid_shape_file: |
|
- exp/asr_stats_raw_bpe1000_sp/valid/speech_shape |
|
- exp/asr_stats_raw_bpe1000_sp/valid/text_shape.bpe |
|
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 |
|
train_data_path_and_name_and_type: |
|
- - dump/raw/train_sp/wav.scp |
|
- speech |
|
- kaldi_ark |
|
- - dump/raw/train_sp/text |
|
- text |
|
- text |
|
valid_data_path_and_name_and_type: |
|
- - dump/raw/dev/wav.scp |
|
- speech |
|
- kaldi_ark |
|
- - dump/raw/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: 4.0 |
|
scheduler: noamlr |
|
scheduler_conf: |
|
model_size: 256 |
|
warmup_steps: 25000 |
|
token_list: |
|
- <blank> |
|
- <unk> |
|
- ▁que |
|
- s |
|
- ▁no |
|
- ▁y |
|
- ▁de |
|
- ▁a |
|
- ▁sí |
|
- ▁la |
|
- ▁es |
|
- ▁en |
|
- r |
|
- n |
|
- ▁yo |
|
- ▁el |
|
- ▁me |
|
- ▁pero |
|
- ▁lo |
|
- ▁o |
|
- ▁se |
|
- ▁un |
|
- ▁los |
|
- ▁como |
|
- ▁ah |
|
- ▁con |
|
- ▁por |
|
- ▁si |
|
- ▁eh |
|
- ▁eso |
|
- ▁una |
|
- ▁porque |
|
- a |
|
- ▁mi |
|
- ▁tu |
|
- ▁para |
|
- ▁más |
|
- ▁ya |
|
- ▁te |
|
- do |
|
- ▁bueno |
|
- ▁hay |
|
- l |
|
- ▁ajá |
|
- ▁pues |
|
- u |
|
- o |
|
- ▁qué |
|
- e |
|
- c |
|
- ▁le |
|
- ▁entonces |
|
- ▁está |
|
- ra |
|
- da |
|
- ▁así |
|
- ▁muy |
|
- ▁las |
|
- ▁claro |
|
- ▁también |
|
- ndo |
|
- ▁todo |
|
- ▁bien |
|
- ▁uno |
|
- d |
|
- ▁sé |
|
- re |
|
- ▁mhm |
|
- la |
|
- to |
|
- ta |
|
- i |
|
- ▁sea |
|
- b |
|
- t |
|
- ▁ca |
|
- é |
|
- te |
|
- es |
|
- ▁son |
|
- ▁aquí |
|
- ▁al |
|
- mos |
|
- ▁gente |
|
- ▁tiene |
|
- ▁ha |
|
- ▁mucho |
|
- ca |
|
- ▁cuando |
|
- ▁oh |
|
- ▁su |
|
- ▁re |
|
- g |
|
- ▁mm |
|
- ti |
|
- ba |
|
- ▁este |
|
- ▁creo |
|
- ▁va |
|
- v |
|
- lo |
|
- ó |
|
- le |
|
- y |
|
- ▁cómo |
|
- í |
|
- ci |
|
- p |
|
- na |
|
- ▁del |
|
- ce |
|
- ▁verdad |
|
- ro |
|
- ▁tengo |
|
- er |
|
- ▁ellos |
|
- ▁era |
|
- ▁ahí |
|
- ▁él |
|
- ▁estoy |
|
- ▁he |
|
- ▁ahora |
|
- m |
|
- ▁ |
|
- f |
|
- ma |
|
- ▁llama |
|
- ▁ma |
|
- ▁cosas |
|
- ri |
|
- ▁años |
|
- en |
|
- ▁hace |
|
- ▁co |
|
- co |
|
- ción |
|
- h |
|
- á |
|
- de |
|
- ▁e |
|
- ▁nada |
|
- ▁casa |
|
- 'no' |
|
- ▁algo |
|
- ▁pa |
|
- ▁estaba |
|
- nta |
|
- ga |
|
- ar |
|
- ▁in |
|
- ▁están |
|
- ▁gusta |
|
- si |
|
- ▁di |
|
- ▁cu |
|
- ▁dos |
|
- mente |
|
- ▁vi |
|
- ▁veces |
|
- ▁uh |
|
- so |
|
- ▁ella |
|
- ▁tienen |
|
- ▁ese |
|
- di |
|
- ▁fue |
|
- ▁hacer |
|
- se |
|
- ▁siempre |
|
- ▁so |
|
- li |
|
- ▁dice |
|
- go |
|
- ▁vez |
|
- ▁soy |
|
- ▁persona |
|
- ▁ba |
|
- ▁acá |
|
- ▁digo |
|
- ía |
|
- ▁ejemplo |
|
- ▁música |
|
- ▁usted |
|
- ron |
|
- ▁ver |
|
- cu |
|
- ▁ve |
|
- ▁ni |
|
- ▁mucha |
|
- sa |
|
- ▁esa |
|
- pe |
|
- ría |
|
- ve |
|
- ▁ser |
|
- ▁okay |
|
- ▁pasa |
|
- z |
|
- ▁puede |
|
- an |
|
- ▁cosa |
|
- ▁da |
|
- ▁otra |
|
- me |
|
- ▁ay |
|
- ▁habla |
|
- al |
|
- ▁sabes |
|
- ▁después |
|
- ja |
|
- ▁tiempo |
|
- nd |
|
- ne |
|
- ado |
|
- mi |
|
- pa |
|
- ▁po |
|
- ▁d |
|
- ▁ju |
|
- ▁i |
|
- ▁otro |
|
- ▁allá |
|
- ▁m |
|
- ica |
|
- ▁estados |
|
- tu |
|
- ▁todos |
|
- nte |
|
- iendo |
|
- va |
|
- ▁donde |
|
- ▁tan |
|
- ▁mismo |
|
- ▁esta |
|
- ▁t |
|
- ▁mo |
|
- ▁ir |
|
- ▁unidos |
|
- ▁trabajo |
|
- ▁poco |
|
- tos |
|
- cho |
|
- ▁menos |
|
- ▁ven |
|
- ▁tenía |
|
- ▁c |
|
- ▁personas |
|
- ▁tener |
|
- za |
|
- ▁mira |
|
- ▁dónde |
|
- mo |
|
- ▁pro |
|
- ▁mejor |
|
- ▁li |
|
- ▁nunca |
|
- ▁decir |
|
- ▁estás |
|
- ▁l |
|
- ▁tra |
|
- ▁ciudad |
|
- ▁per |
|
- rio |
|
- ▁pre |
|
- ▁voy |
|
- ▁exacto |
|
- ▁tienes |
|
- ▁había |
|
- ▁f |
|
- ▁sabe |
|
- tra |
|
- sta |
|
- ▁muchos |
|
- idad |
|
- ▁país |
|
- ▁p |
|
- ▁mu |
|
- ▁hasta |
|
- ▁parte |
|
- ▁igual |
|
- ▁muchas |
|
- ▁día |
|
- mp |
|
- ▁comp |
|
- ▁parece |
|
- ido |
|
- ciones |
|
- ▁pu |
|
- ▁esto |
|
- ▁nueva |
|
- ▁nosotros |
|
- jo |
|
- ▁ex |
|
- ▁problema |
|
- ▁ro |
|
- ▁pe |
|
- ▁tema |
|
- cia |
|
- k |
|
- ble |
|
- ▁do |
|
- ▁tres |
|
- ▁van |
|
- ▁sa |
|
- ▁um |
|
- ▁hm |
|
- ▁estar |
|
- ▁sal |
|
- ▁s |
|
- fi |
|
- je |
|
- ▁hola |
|
- ▁york |
|
- ▁tanto |
|
- os |
|
- ge |
|
- ▁vida |
|
- ▁familia |
|
- ▁ra |
|
- ▁les |
|
- das |
|
- ▁teléfono |
|
- ie |
|
- ▁mundo |
|
- ▁hu |
|
- ▁digamos |
|
- ▁quiere |
|
- nos |
|
- ▁bastante |
|
- ste |
|
- j |
|
- qui |
|
- les |
|
- ▁hablar |
|
- res |
|
- ▁b |
|
- ▁u |
|
- ▁español |
|
- ▁tal |
|
- ▁dios |
|
- che |
|
- ▁han |
|
- ▁dinero |
|
- pi |
|
- ni |
|
- ▁difícil |
|
- st |
|
- ▁v |
|
- ▁gra |
|
- dos |
|
- ue |
|
- ▁chi |
|
- ▁em |
|
- ▁dicen |
|
- ▁antes |
|
- tas |
|
- era |
|
- ▁año |
|
- ▁vive |
|
- ▁cierto |
|
- ia |
|
- rá |
|
- ria |
|
- po |
|
- nt |
|
- ▁religión |
|
- que |
|
- ▁ci |
|
- ▁cinco |
|
- ▁ne |
|
- ió |
|
- ▁cre |
|
- ente |
|
- ñ |
|
- ir |
|
- ▁car |
|
- ▁to |
|
- bo |
|
- ▁casi |
|
- men |
|
- ▁niños |
|
- ▁ti |
|
- bu |
|
- w |
|
- il |
|
- ▁cada |
|
- ieron |
|
- cha |
|
- ▁dije |
|
- x |
|
- ▁pienso |
|
- ▁g |
|
- ▁llega |
|
- ento |
|
- ▁grande |
|
- ▁todavía |
|
- ▁realmente |
|
- ▁alguna |
|
- ▁todas |
|
- ▁mar |
|
- cio |
|
- ▁acuerdo |
|
- mb |
|
- ▁segu |
|
- rse |
|
- ▁mal |
|
- ▁estamos |
|
- ▁tipo |
|
- be |
|
- ▁pone |
|
- ▁eres |
|
- ▁sin |
|
- ▁tenemos |
|
- encia |
|
- ▁alguien |
|
- cto |
|
- tar |
|
- ▁vos |
|
- ▁fi |
|
- ▁haciendo |
|
- ▁quién |
|
- ▁toda |
|
- ▁viene |
|
- io |
|
- ura |
|
- ▁pen |
|
- ▁hombre |
|
- ▁hacen |
|
- ▁hablando |
|
- ▁ayuda |
|
- ▁hi |
|
- ▁trata |
|
- ▁hoy |
|
- ito |
|
- ten |
|
- ▁na |
|
- ▁exactamente |
|
- ▁escucha |
|
- ver |
|
- un |
|
- ▁conoce |
|
- gue |
|
- ño |
|
- ▁filadelfia |
|
- ita |
|
- ▁poder |
|
- ▁fa |
|
- án |
|
- ▁puedo |
|
- ▁lugar |
|
- ▁vamos |
|
- ▁yeah |
|
- ú |
|
- ▁desde |
|
- ▁pi |
|
- lla |
|
- ▁hora |
|
- lu |
|
- ▁otros |
|
- ▁méxico |
|
- ▁internet |
|
- ▁res |
|
- ▁solamente |
|
- ban |
|
- ▁usa |
|
- ▁vas |
|
- ▁fui |
|
- ▁estado |
|
- ▁lleva |
|
- ▁mil |
|
- ▁solo |
|
- ▁entre |
|
- ▁wow |
|
- ▁éste |
|
- ch |
|
- den |
|
- ▁frío |
|
- ▁cree |
|
- ▁caso |
|
- ▁estudia |
|
- ▁am |
|
- ▁busca |
|
- ▁dis |
|
- ▁trabaja |
|
- ▁fe |
|
- ▁bo |
|
- ▁hecho |
|
- ▁pueden |
|
- ▁poquito |
|
- ter |
|
- ▁vivir |
|
- ▁amigo |
|
- ▁cuánto |
|
- ▁ga |
|
- ▁esas |
|
- ul |
|
- ▁tampoco |
|
- ▁hijos |
|
- ▁unos |
|
- el |
|
- ▁cuatro |
|
- ▁sus |
|
- ▁com |
|
- ▁aunque |
|
- ▁seguro |
|
- ▁ce |
|
- ▁forma |
|
- ▁debe |
|
- los |
|
- ▁ta |
|
- cion |
|
- ones |
|
- ▁puedes |
|
- ▁mamá |
|
- ▁cuenta |
|
- ▁mis |
|
- ▁diferente |
|
- ▁quiero |
|
- ▁ho |
|
- ▁vivo |
|
- ▁celular |
|
- ero |
|
- ▁universidad |
|
- ▁be |
|
- ▁misma |
|
- ▁deja |
|
- ▁cuál |
|
- ▁inglés |
|
- ▁nombre |
|
- dia |
|
- ▁paga |
|
- ▁ahorita |
|
- ▁cambia |
|
- gra |
|
- ▁dan |
|
- ▁allí |
|
- ▁rico |
|
- ▁puerto |
|
- ▁buenas |
|
- ▁manera |
|
- ▁cri |
|
- ▁días |
|
- ▁ésta |
|
- ▁cualquier |
|
- ▁países |
|
- ing |
|
- tica |
|
- ina |
|
- ▁buena |
|
- ▁nadie |
|
- ▁decía |
|
- ▁piensa |
|
- ▁sobre |
|
- ▁esposo |
|
- ▁qui |
|
- ▁chile |
|
- tro |
|
- ▁toma |
|
- ▁dijo |
|
- ▁quieren |
|
- ▁película |
|
- ▁semana |
|
- ▁sistema |
|
- ▁come |
|
- ▁mujer |
|
- ▁veo |
|
- ▁n |
|
- ida |
|
- ▁otras |
|
- ▁medio |
|
- ón |
|
- ▁diez |
|
- ▁cerca |
|
- ▁iba |
|
- ico |
|
- gan |
|
- ▁necesita |
|
- zo |
|
- ▁oye |
|
- ▁san |
|
- ▁bu |
|
- ▁entiendes |
|
- tó |
|
- rme |
|
- ▁sería |
|
- ▁argentina |
|
- ▁momento |
|
- miento |
|
- ▁know |
|
- ▁seis |
|
- ▁fo |
|
- ▁toca |
|
- ▁manda |
|
- ▁w |
|
- ▁iglesia |
|
- ▁dólares |
|
- ▁ja |
|
- ▁diferentes |
|
- ista |
|
- ▁escuela |
|
- ▁fácil |
|
- ▁sale |
|
- ▁interesante |
|
- ▁padre |
|
- ▁gana |
|
- ▁inter |
|
- ▁papá |
|
- mina |
|
- ▁pregunta |
|
- iente |
|
- ▁realidad |
|
- ▁conozco |
|
- ▁dar |
|
- sión |
|
- ▁tenido |
|
- ▁trabajar |
|
- ▁pareja |
|
- ▁gu |
|
- ▁mío |
|
- ▁hijo |
|
- ig |
|
- ▁vivi |
|
- ▁computadora |
|
- ▁visto |
|
- ▁importante |
|
- ▁pasado |
|
- ▁vol |
|
- ▁tenga |
|
- ho |
|
- ▁pagar |
|
- ▁latino |
|
- ▁corre |
|
- ▁haber |
|
- ▁televisión |
|
- ▁luego |
|
- ▁relación |
|
- ▁señor |
|
- ▁tanta |
|
- ▁mujeres |
|
- iza |
|
- ▁treinta |
|
- ▁idea |
|
- ▁salir |
|
- ▁americano |
|
- ▁encanta |
|
- ▁meses |
|
- ▁pasó |
|
- ▁programa |
|
- ▁algún |
|
- ▁pri |
|
- ▁estuve |
|
- ▁comprar |
|
- ▁contra |
|
- ▁bonito |
|
- ▁colombia |
|
- ▁compra |
|
- ▁super |
|
- ▁hacía |
|
- ▁imp |
|
- ▁cultura |
|
- ▁fíjate |
|
- ▁sino |
|
- ▁poner |
|
- ▁fuera |
|
- ▁ri |
|
- ▁veinte |
|
- ▁buen |
|
- ▁único |
|
- ▁entiendo |
|
- ▁depende |
|
- ▁fu |
|
- ▁españa |
|
- ▁quizás |
|
- ▁esté |
|
- ▁gracias |
|
- ▁hija |
|
- tico |
|
- ▁imagino |
|
- q |
|
- ▁quiera |
|
- ▁comuni |
|
- ▁espera |
|
- ▁go |
|
- ▁primera |
|
- ▁clase |
|
- ▁general |
|
- ▁diciendo |
|
- ▁carro |
|
- ▁anda |
|
- ▁somos |
|
- ▁sabía |
|
- ▁amiga |
|
- ▁vaya |
|
- ▁compañía |
|
- ▁siete |
|
- ▁viste |
|
- ▁canadá |
|
- ▁cuanto |
|
- ▁empeza |
|
- ▁mayor |
|
- ▁lleg |
|
- ▁ido |
|
- ▁malo |
|
- ▁debería |
|
- ▁gobierno |
|
- ▁edad |
|
- ▁situación |
|
- ▁trabajando |
|
- tivo |
|
- ▁calle |
|
- ▁veinti |
|
- ▁mayoría |
|
- ▁plan |
|
- ▁viviendo |
|
- ▁termina |
|
- ▁llamo |
|
- ▁viaja |
|
- ▁social |
|
- ▁jo |
|
- ▁ciento |
|
- ▁joven |
|
- ▁estudio |
|
- ▁hablo |
|
- ▁empieza |
|
- ▁podía |
|
- ▁baila |
|
- ▁punto |
|
- ▁matrimonio |
|
- ▁primero |
|
- ▁entiende |
|
- ▁perdón |
|
- ▁niña |
|
- ▁pobre |
|
- fect |
|
- ▁hispano |
|
- ▁auto |
|
- ▁importa |
|
- ▁tarde |
|
- ▁vivía |
|
- ▁gustaría |
|
- ▁diferencia |
|
- ▁pueda |
|
- ▁experiencia |
|
- ▁ángeles |
|
- ▁pie |
|
- ▁oportunidad |
|
- ▁mañana |
|
- ▁nuevo |
|
- ▁ningún |
|
- ▁k |
|
- ▁razón |
|
- ▁minutos |
|
- vis |
|
- ▁además |
|
- ▁cha |
|
- ▁nueve |
|
- ▁comercial |
|
- ▁demasiado |
|
- ▁encontrar |
|
- port |
|
- ▁sentido |
|
- ▁número |
|
- ▁política |
|
- ▁niño |
|
- ▁grupo |
|
- ▁pensar |
|
- ▁hermano |
|
- ísimo |
|
- ▁raza |
|
- ▁afuera |
|
- ▁quince |
|
- ▁sitio |
|
- ▁policía |
|
- ▁gusto |
|
- ▁fuerte |
|
- ▁miami |
|
- ▁palabra |
|
- ▁montón |
|
- ▁cincuenta |
|
- ▁falta |
|
- ▁recuerdo |
|
- ▁visita |
|
- ▁normal |
|
- ▁especialmente |
|
- ▁hizo |
|
- ▁salud |
|
- ▁partido |
|
- ▁plata |
|
- ▁venezuela |
|
- ▁ru |
|
- ▁novia |
|
- ▁cierta |
|
- ▁educa |
|
- ▁área |
|
- ▁maneja |
|
- ▁quien |
|
- ▁acostumbra |
|
- ▁conocí |
|
- ▁doctor |
|
- ▁inmigrante |
|
- ▁básicamente |
|
- ▁mexicano |
|
- ▁comida |
|
- ▁algunos |
|
- ▁enseña |
|
- ▁cuarenta |
|
- ▁supuesto |
|
- ▁panamá |
|
- ▁religiones |
|
- ▁cuestión |
|
- ▁bi |
|
- ▁final |
|
- ▁encuentro |
|
- ▁llevo |
|
- ▁tenés |
|
- ▁hermana |
|
- ▁papel |
|
- ▁existe |
|
- ▁aprende |
|
- ▁novio |
|
- ▁encontr |
|
- ▁cambio |
|
- ▁negocio |
|
- ▁atrás |
|
- ▁podría |
|
- ▁miedo |
|
- ismo |
|
- ▁increíble |
|
- ▁pongo |
|
- ▁aparte |
|
- ▁osea |
|
- ▁médico |
|
- ▁acento |
|
- ▁terrible |
|
- ▁enferm |
|
- ▁hablé |
|
- ▁regresa |
|
- ▁texas |
|
- ▁jurado |
|
- ▁última |
|
- ▁peor |
|
- ▁estuvo |
|
- ▁dentro |
|
- ▁color |
|
- ▁viví |
|
- ▁right |
|
- ▁chicago |
|
- ▁servicio |
|
- ▁interesa |
|
- ▁muchísimo |
|
- ▁email |
|
- ▁escucho |
|
- ▁pronto |
|
- ▁homosexual |
|
- ▁rápido |
|
- ▁esposa |
|
- ▁principio |
|
- ▁llen |
|
- ▁hospital |
|
- ▁imagínate |
|
- ▁peligro |
|
- ▁cuándo |
|
- ▁uhum |
|
- ▁apartamento |
|
- ▁funciona |
|
- ▁historia |
|
- ▁tecnología |
|
- ▁control |
|
- ▁ninguna |
|
- ▁juntos |
|
- ▁encuentra |
|
- ▁horrible |
|
- ▁centro |
|
- ▁atención |
|
- ▁hubiera |
|
- ▁totalmente |
|
- ▁california |
|
- ▁católica |
|
- ▁molesta |
|
- ▁gustó |
|
- ▁información |
|
- ▁méjico |
|
- ▁suerte |
|
- ▁argentino |
|
- ▁divi |
|
- ▁florida |
|
- ▁guerra |
|
- ▁aires |
|
- ▁nieve |
|
- ▁obviamente |
|
- ▁pelea |
|
- ▁nuestro |
|
- ▁simplemente |
|
- ▁pequeño |
|
- ▁clima |
|
- ▁europa |
|
- ▁imagina |
|
- ▁arriba |
|
- ▁leyes |
|
- ▁playa |
|
- ▁violencia |
|
- ▁conversa |
|
- ▁fiesta |
|
- ▁tranquilo |
|
- ▁acepta |
|
- ▁último |
|
- ▁única |
|
- ▁definitivamente |
|
- ▁incluso |
|
- ▁idioma |
|
- ▁favor |
|
- ▁blanco |
|
- ▁presidente |
|
- ▁invierno |
|
- ▁separa |
|
- ivo |
|
- ▁primer |
|
- ▁nuestra |
|
- ▁bonita |
|
- ▁culpa |
|
- ▁vota |
|
- ▁entendí |
|
- ▁madre |
|
- ▁conocido |
|
- ▁arregl |
|
- ▁acerca |
|
- ▁washington |
|
- ▁radio |
|
- ▁opina |
|
- ▁contigo |
|
- ▁podemos |
|
- ▁pensando |
|
- ▁duro |
|
- ▁conmigo |
|
- ▁verano |
|
- '0' |
|
- ▁negro |
|
- ▁mientras |
|
- ▁nací |
|
- ▁toronto |
|
- ▁recibi |
|
- ▁hicieron |
|
- ▁boston |
|
- ▁campo |
|
- ▁repente |
|
- ▁cocina |
|
- ▁cuesta |
|
- ▁conseguir |
|
- ▁jóvenes |
|
- ▁olvida |
|
- ▁ochenta |
|
- ▁nivel |
|
- ▁sociedad |
|
- ▁chiquito |
|
- ▁guatemala |
|
- ▁político |
|
- ▁supongo |
|
- ▁empezó |
|
- ▁época |
|
- ▁siquiera |
|
- ▁agarra |
|
- ▁católico |
|
- ▁pennsylvania |
|
- ▁medicina |
|
- ▁entender |
|
- ▁italia |
|
- ▁especial |
|
- ▁atlanta |
|
- ▁navidad |
|
- ▁cantidad |
|
- ▁domingo |
|
- ▁cristiano |
|
- ▁opinión |
|
- ▁crédito |
|
- ▁noticias |
|
- ▁houston |
|
- ▁preocupa |
|
- ▁mensaje |
|
- ▁américa |
|
- ▁perfecto |
|
- ▁dijiste |
|
- '1' |
|
- '2' |
|
- '5' |
|
- _ |
|
- '-' |
|
- '3' |
|
- '6' |
|
- '4' |
|
- '9' |
|
- '8' |
|
- '7' |
|
- A |
|
- B |
|
- ì |
|
- à |
|
- ç |
|
- è |
|
- ü |
|
- <sos/eos> |
|
init: null |
|
input_size: null |
|
ctc_conf: |
|
dropout_rate: 0.0 |
|
ctc_type: builtin |
|
reduce: true |
|
ignore_nan_grad: null |
|
zero_infinity: true |
|
joint_net_conf: null |
|
use_preprocessor: true |
|
token_type: bpe |
|
bpemodel: data/token_list/bpe_unigram1000/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' |
|
short_noise_thres: 0.5 |
|
aux_ctc_tasks: [] |
|
frontend: default |
|
frontend_conf: |
|
n_fft: 512 |
|
win_length: 400 |
|
hop_length: 160 |
|
fs: 8k |
|
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_bpe1000_sp/train/feats_stats.npz |
|
model: espnet |
|
model_conf: |
|
ctc_weight: 0.3 |
|
lsm_weight: 0.1 |
|
length_normalized_loss: false |
|
preencoder: null |
|
preencoder_conf: {} |
|
encoder: e_branchformer |
|
encoder_conf: |
|
output_size: 256 |
|
attention_heads: 4 |
|
attention_layer_type: rel_selfattn |
|
pos_enc_layer_type: rel_pos |
|
rel_pos_type: latest |
|
cgmlp_linear_units: 1024 |
|
cgmlp_conv_kernel: 31 |
|
use_linear_after_conv: false |
|
gate_activation: identity |
|
num_blocks: 16 |
|
dropout_rate: 0.1 |
|
positional_dropout_rate: 0.1 |
|
attention_dropout_rate: 0.1 |
|
input_layer: conv2d |
|
layer_drop_rate: 0.0 |
|
linear_units: 1024 |
|
positionwise_layer_type: linear |
|
use_ffn: true |
|
macaron_ffn: true |
|
merge_conv_kernel: 31 |
|
postencoder: null |
|
postencoder_conf: {} |
|
decoder: transformer |
|
decoder_conf: |
|
input_layer: embed |
|
num_blocks: 6 |
|
linear_units: 2048 |
|
dropout_rate: 0.1 |
|
preprocessor: default |
|
preprocessor_conf: {} |
|
required: |
|
- output_dir |
|
- token_list |
|
version: '202301' |
|
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} |
|
} |
|
``` |
|
|
