config: conf/train_asr.yaml
print_config: false
log_level: INFO
dry_run: false
iterator_type: sequence
output_dir: exp/asr_train_asr_raw_bpe1000
ngpu: 1
seed: 0
num_workers: 1
num_att_plot: 3
dist_backend: nccl
dist_init_method: env://
dist_world_size: 2
dist_rank: 0
local_rank: 0
dist_master_addr: localhost
dist_master_port: 44079
dist_launcher: null
multiprocessing_distributed: true
unused_parameters: false
sharded_ddp: false
cudnn_enabled: true
cudnn_benchmark: false
cudnn_deterministic: true
collect_stats: false
write_collected_feats: false
max_epoch: 200
patience: 20
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_bpe1000/train/speech_shape
- exp/asr_stats_raw_bpe1000/train/text_shape.bpe
valid_shape_file:
- exp/asr_stats_raw_bpe1000/valid/speech_shape
- exp/asr_stats_raw_bpe1000/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:
-   - dump/raw/sunda_train/wav.scp
    - speech
    - sound
-   - dump/raw/sunda_train/text
    - text
    - text
valid_data_path_and_name_and_type:
-   - dump/raw/sunda_dev/wav.scp
    - speech
    - sound
-   - dump/raw/sunda_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: 10.0
scheduler: noamlr
scheduler_conf:
    warmup_steps: 25000
token_list:
- <blank>
- <unk>
- ▁
- ▁DI
- KEUR
- ▁KU
- ▁JEUNG
- ▁OL
- ▁NINGALI
- NCARA
- ▁WARTAWAN
- OHOK
- WAWA
- POTO
- A
- I
- E
- Y
- S
- O
- AN
- T
- NA
- N
- R
- RA
- L
- ER
- H
- D
- ▁KA
- 'ON'
- U
- TA
- IN
- ▁K
- ▁S
- P
- MA
- G
- K
- AR
- LA
- RI
- NG
- M
- EL
- ▁M
- ▁D
- DI
- LE
- B
- IE
- ▁P
- Z
- W
- WI
- ▁B
- ▁MAR
- EN
- AD
- ▁T
- .
- SON
- ▁F
- AS
- IS
- UR
- KA
- ▁LI
- MAN
- ING
- UNG
- ▁G
- ▁AL
- V
- TON
- UN
- WA
- ▁AYA
- ▁POE
- F
- ▁TI
- ▁DINA
- ▁NI
- TH
- RE
- ▁O
- ▁ARI
- ▁HA
- CA
- NE
- CK
- ORAN
- ▁JA
- ▁ME
- TO
- AH
- ▁SI
- US
- ▁C
- ▁SO
- UL
- LO
- DEWI
- ES
- ▁BA
- ▁Z
- ▁MI
- AL
- ▁RO
- ED
- ▁SA
- KY
- ▁PA
- ▁SU
- EK
- ▁BE
- SA
- TYA
- ▁CA
- IKA
- ▁BEN
- NDA
- CE
- ▁JO
- NATA
- ▁H
- IA
- ▁RU
- INA
- ▁CHRIS
- ▁SAND
- ▁THE
- IM
- BERT
- ▁SAM
- YA
- C
- HA
- ▁HO
- ▁LEE
- KU
- GO
- ▁DON
- ▁IR
- ATI
- UK
- ▁W
- GER
- ZI
- ▁JOHN
- ▁IS
- ANG
- ▁KE
- LEY
- ▁RE
- ▁SE
- ▁BO
- IF
- ▁MUHAM
- MY
- ▁JACK
- LY
- DO
- KI
- ▁KAT
- ▁RAM
- KH
- IC
- ▁L
- ▁JI
- AM
- ▁JESSI
- ▁BELL
- ▁MO
- AVID
- RO
- HAM
- EM
- OR
- JA
- ▁HE
- SARI
- MAD
- ID
- '-'
- LAN
- OT
- RAH
- ASA
- ▁ELIZABETH
- INE
- ▁TOM
- ▁LE
- ▁LA
- ▁REZ
- ▁AND
- ▁JENN
- UM
- ▁MARIA
- ▁BON
- ▁CHO
- FA
- YONG
- ▁TITI
- HEN
- ▁BR
- EB
- ▁ANU
- ISA
- ▁PE
- ▁DO
- ▁TEU
- TEN
- ▁MIR
- ▁NU
- WAN
- ▁RI
- EZ
- ▁VIN
- MAL
- OS
- OOR
- RAN
- ▁WIN
- BING
- ▁YU
- ▁GI
- ▁AHM
- ANTI
- ▁LU
- ALI
- ▁NO
- ▁BILL
- ▁YO
- J
- SHA
- ▁ADAM
- LIE
- ▁SOPHI
- CKY
- ▁JON
- ▁MARTIN
- ▁KHAN
- DER
- ▁KAP
- ▁JOE
- ▁WON
- ▁JAME
- ▁WILLIAM
- ▁J
- ▁KAR
- ▁AR
- MITH
- ▁ADI
- UA
- ▁CHANDRA
- ▁RIZ
- ▁ASHLE
- ▁RICHARD
- ▁PUTRI
- ▁JOSH
- ▁HARRIS
- ▁GUN
- RINA
- ▁INDRA
- ARA
- ▁LAT
- ▁EM
- ▁AYU
- ▁VAN
- WATI
- CHAN
- SYA
- ▁ALI
- CHEL
- ▁PAN
- ALL
- BU
- ▁CH
- AWA
- ▁ABD
- NS
- ERRY
- AEL
- ▁DAR
- ▁JU
- NY
- ▁SHAH
- ELLY
- HU
- ▁MICHEL
- ▁CAR
- RIO
- ULIN
- ▁DAHLIA
- ▁KANG
- ▁MIN
- ▁LOBA
- ▁YUKI
- ▁EVA
- ▁WILL
- INDAH
- ▁JUD
- NTY
- ▁CIT
- ▁JUN
- ▁MC
- ▁JUSTI
- PERT
- ▁CI
- ▁US
- ▁YUN
- ▁SIMP
- NIE
- DAD
- ▁MEL
- ▁SHERI
- ▁PAUL
- RYAN
- ▁RAN
- ▁VI
- HAKIM
- ▁FAT
- ALLEN
- UNGAN
- ▁MARC
- MES
- ▁CRIST
- ▁MARK
- ▁DANIEL
- ▁ROSS
- FAN
- ▁REEVE
- ▁JULIA
- ZHA
- ONEY
- ▁KANTO
- ▁KATE
- ▁JIM
- ▁LAU
- ▁WOOD
- BAMA
- LOR
- LUM
- GHAN
- ▁TRI
- ▁MICH
- ▁DU
- RTIKA
- ▁RICK
- ESSA
- TIK
- EVIN
- ▁TAY
- ANDAR
- ▁SHI
- ELA
- ▁GAD
- ▁WIB
- ▁JE
- CHOLAS
- YAM
- ▁HARU
- ▁TUR
- ▁CHARL
- ▁SHAR
- JAYA
- ▁HYUN
- X
- ▁SYAH
- ▁LANCE
- LVIN
- UKI
- ▁PAPA
- ▁DARA
- AUDY
- ▁SHA
- ▁SIT
- ARAH
- SEA
- ▁BUL
- ▁CO
- BANG
- KEUN
- LINA
- RYA
- ▁NGAR
- RIANA
- ▁GEDE
- AWAN
- ▁PANG
- ▁MELL
- AKE
- STIAN
- NDRA
- ▁CAFE
- DAY
- ▁BAD
- ▁LAUR
- ▁KRIS
- ▁REBO
- ▁NEP
- ▁ROBB
- ▁STONE
- BAH
- WARD
- ▁FAR
- LIA
- ▁SORA
- PATI
- ▁RATU
- ▁KAMAR
- ▁SUN
- ▁MEDIA
- EPP
- ▁AGUS
- ▁MASA
- NDY
- ▁LAM
- NDES
- TINA
- ▁DEV
- ISH
- ▁GREEN
- ORO
- BALE
- ▁SURYA
- ▁PARK
- RAHMAN
- IPUT
- ▁GINANJAR
- REW
- ▁VICTORIA
- ▁JANET
- ▁BAY
- OSTER
- ▁STE
- ▁RITA
- ▁ARM
- ▁MAHER
- ▁CARE
- MELA
- WART
- ALIK
- ▁ANGEL
- STRONG
- ▁CRUISE
- ▁DANI
- ▁CLIN
- ▁AYEUNA
- ▁SONG
- ▁FACHR
- GITA
- RANDA
- ▁MAU
- KITA
- ICIA
- ▁FEB
- ▁ANT
- VIA
- ▁TURNE
- PORT
- ▁RAS
- LIU
- ▁OCTA
- ▁KHALI
- OPH
- ▁STEP
- ▁WEST
- ▁SEUNG
- HIM
- ▁FERNAND
- LESTARI
- ▁RY
- SCAR
- REGAR
- AMIR
- SAPUTRA
- ▁AGNES
- ▁MULYA
- HONY
- AFFLE
- ▁PERSI
- ▁CHE
- OLIVIA
- ▁SALMA
- ▁RONAL
- HOPE
- HUN
- ▁ROSE
- ILARI
- ▁KRISTEN
- ▁NOVITA
- ▁KAKA
- AMIL
- ▁RON
- ▁CITA
- ALBAR
- ▁VIR
- ▁FENI
- IOR
- ▁LED
- ▁ELVI
- ▁YOON
- ▁RIDWAN
- ▁RAF
- ▁ASTRI
- ▁DUA
- ▁MILL
- ▁DES
- IEF
- ▁TAM
- ▁ROY
- ▁KOES
- SUMA
- ▁CHIN
- ▁ROD
- ▁PIT
- KARNA
- LANK
- ▁YAYAN
- ▁BIG
- ▁ANIS
- OHANG
- QI
- ▁ALAM
- ▁MEG
- MAHE
- OWO
- ▁DOMIN
- ▁CHER
- IMBERL
- INTAN
- MORE
- ▁SUKA
- ▁NINI
- ▁STEFAN
- ▁CASE
- NAJ
- SEUEUR
- STER
- ▁JORDAN
- ▁ARIF
- ▁KIRAN
- YAKI
- ETTI
- ▁YUNG
- ASIH
- ▁SHEI
- QUE
- ▁STEV
- ▁VER
- MAJA
- ▁WAHYU
- ▁MICHE
- ▁BIMB
- HAB
- ▁BAHA
- ▁CAME
- ▁ALEXAND
- ▁ZAC
- ▁FRED
- ▁PIER
- ▁PINK
- ▁THOM
- ▁DJ
- ▁EMIL
- GIE
- ▁NICO
- ▁RUS
- ▁INDONESIA
- ▁KURNIA
- ▁FITRI
- IKO
- YANA
- ▁BOW
- LETT
- ▁HOL
- ▁MONA
- ▁ANDY
- GHAZ
- ▁MAG
- ▁NAD
- LANA
- ESIH
- ASMI
- ▁ARMAND
- ▁ADE
- DJA
- OVA
- AURE
- KUMAH
- ▁DANU
- ▁JERE
- ▁KIM
- ▁HAZ
- ▁GAL
- DANE
- ▁LEV
- ▁AMIN
- IBRA
- ▁MELAN
- ▁LOUIS
- ▁PURNAMA
- CENT
- ▁BRAMA
- ▁DANANG
- ▁DEVI
- ▁RANGGA
- ▁IQBA
- ▁POPP
- AYN
- HETT
- ▁DIMAS
- ▁MAND
- ▁NGAL
- ▁AGU
- ▁IVAN
- ▁FANN
- ▁IB
- ▁CLARK
- ▁HERMAN
- ▁ANA
- RELL
- LIAM
- LENN
- USSY
- ▁AHAD
- ▁RIRI
- OVI
- NTEN
- BRATA
- RID
- ▁ANNI
- ▁HERLIN
- ▁SULISTY
- SUHU
- ▁ANDRE
- ▁SENEN
- ▁NAND
- NAGARA
- ▁ANGGA
- ▁KRIST
- ACH
- PIRANG
- NDER
- DARI
- JUNG
- STADI
- ▁SALASA
- RLAND
- ▁NGOPI
- ▁JUMAAH
- ▁MANG
- RANI
- DHI
- ▁RAMADHAN
- ▁KIR
- ▁RATUS
- ▁AISH
- DIANA
- ▁MINGGON
- ▁KEMIS
- ▁LEO
- ASHI
- ▁SHAN
- ▁BANDUNG
- ▁IWA
- ▁PANTAI
- WEN
- ▁AWAK
- ▁PRESIDEN
- ▁NIR
- ▁MANCA
- ▁SAPTU
- CHA
- ▁KARD
- NTUH
- ▁MUSE
- STASI
- KMAN
- ▁FANS
- ▁ANYAR
- ▁ABI
- NGIR
- ▁DONGKAP
- '5'
- NGSI
- Q
- ▁PROPERT
- ▁NYIA
- NGGIH
- ▁DIDAMEL
- ▁NASIONAL
- OLI
- ▁KAAYA
- ▁KYL
- ▁UJANG
- ▁PENGG
- ▁POSTER
- ▁IEU
- IMAH
- ANDRY
- ▁KAGIATAN
- ▁KABUL
- ARIM
- ▁RUPIAH
- CAGE
- ▁MUSI
- ▁PART
- IBURAN
- ▁GAGA
- SUF
- ▁EFFE
- GEND
- ▁BLACK
- ▁ALBA
- ▁KUNC
- ▁VALEN
- IJAH
- ▁ANTARA
- OODY
- JOL
- ▁FOO
- ▁POLAND
- ▁METAL
- HYUN
- ADONA
- ▁ALEXA
- ARTI
- ▁BUSH
- ▁NORT
- LICA
- ▁GAMBAR
- ▁DOWN
- ▁QUEEN
- ▁STREE
- ▁CUT
- ▁BROK
- BEG
- ▁STAN
- ▁QUI
- GBA
- LEVI
- RIGHT
- ▁SEBA
- ▁WIZ
- IPAN
- ▁SUMBANG
- ▁ROLL
- ▁KARNO
- ESTU
- ▁ADJ
- ▁BIL
- AVO
- ▁NGADA
- ▁KOKO
- ▁NABI
- ▁GOES
- ▁FIGHT
- ▁STILL
- ▁PATT
- RRES
- ERENA
- ILANG
- ISTY
- ATUM
- ▁GUARDI
- ERJUN
- ▁NGAGE
- ▁LOREN
- ETRAL
- BIR
- FRON
- ▁SAHI
- ▁MISS
- ▁GRAND
- ▁GIGI
- ▁HYD
- ▁APPLE
- UYAN
- ▁WINTER
- ▁PETRA
- MAJU
- ▁BATU
- ▁MIHAR
- ▁BIDE
- ▁CECEP
- DIANTO
- DINI
- ▁AFGAN
- ▁WALK
- ▁CASI
- ▁DAMON
- ▁GARE
- ▁GAWE
- ▁MERC
- ▁TRAIN
- ▁TULU
- ▁KEYS
- ▁NGALAMA
- ▁BROWN
- ▁MAYORI
- ▁ELLI
- ▁GEUS
- ▁MEUNANG
- ▁KEND
- IKER
- SUT
- RAIG
- ▁SYARIF
- TEMAT
- ORGAN
- COTT
- LLAS
- ▁PANDU
- ▁PANGAB
- ▁WALT
- ▁SUPER
- EXTO
- IPT
- ▁MOSS
- ▁NASSA
- ▁HUG
- NADIA
- JUB
- HANKS
- ▁GRACE
- EATING
- ARDIAN
- ▁WISNU
- ▁ANGKAT
- ▁KOME
- NCER
- BIP
- SLIM
- ▁HYE
- USSE
- RENCE
- LOPE
- AKAMA
- SAHANA
- ▁INTERNET
- ▁MUKTI
- UTS
- ▁TERE
- ▁POP
- ▁MULAN
- ▁BUNGA
- ▁USAHA
- BRUG
- ▁SCR
- ▁WATSON
- ▁SLAV
- PLAY
- ▁RIYAD
- ▁FELT
- RAINI
- ▁MUCH
- AMUNGK
- ▁BANK
- ▁DUNYA
- ▁FALS
- BET
- MUSU
- ▁BLOOM
- ▁HIGH
- ▁LILIS
- ▁KOLO
- ▁BRUNO
- ▁AJI
- COCK
- ▁PRIM
- ▁UNTUNG
- ▁STAT
- ▁HILL
- ▁ROBIN
- ▁GRAVE
- ▁MORRIS
- RAJAT
- ▁GEORGE
- ▁UTAMI
- ▁FORD
- ▁PISAN
- ▁JULI
- ROACH
- HIK
- ▁SAKULIA
- ▁CAKRA
- ▁SOLIHUN
- RUMP
- ▁NANI
- CULES
- HINI
- ▁HYU
- ▁GEORGI
- ▁PUTRA
- ▁JEONG
- VITA
- ▁REST
- ▁PEGG
- LUNT
- ONALD
- ▁BEAT
- ICHO
- AVEN
- ▁SARDI
- OLEH
- PUTIN
- TIANA
- ▁CUCU
- HARGA
- PUSPA
- ▁CROW
- ▁ISLAM
- ▁PERMAN
- ▁NUNU
- ▁BUBBLE
- ▁SUKSES
- ▁SUGAND
- BAGJA
- LESTI
- IFFE
- PHAR
- ▁APRIL
- ▁AGAT
- ▁ARCH
- ▁SHEE
- ▁GILL
- ▁COFF
- STISI
- ▁ENDANG
- INDRO
- OBAN
- GALUNG
- PPIE
- ▁PETE
- ▁BAREN
- BOURNE
- LESIA
- MAIL
- ▁PRINCE
- ▁LANGIT
- SWOR
- SEBAT
- TISS
- UWAI
- ▁WEI
- NIGHT
- SADEW
- DALE
- AMBO
- IGG
- AGASKAR
- ▁HEATH
- RESNA
- HALLE
- AMIT
- ▁BRANCH
- AKOTA
- ▁BARB
- ▁CHEMI
- ▁PRIATNA
- '0'
- ▁COBA
- ▁ESPE
- ITJE
- WULAN
- ▁GUGUM
- NSIL
- UNGSU
- ▁BLUE
- ▁ASEP
- TILES
- UCI
- ITEPU
- ▁PROJECT
- PPOL
- Í
- '1'
- '"'
- '6'
- '2'
- '3'
- '8'
- ''''
- '9'
- '7'
- '4'
- '!'
- È
- ','
- ;
- '&'
- Ä
- Ñ
- Ō
- <sos/eos>
init: chainer
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_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'
frontend: default
frontend_conf:
    fs: 16k
specaug: null
specaug_conf: {}
normalize: global_mvn
normalize_conf:
    stats_file: exp/asr_stats_raw_bpe1000/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: 256
    attention_heads: 4
    attention_dropout_rate: 0.0
decoder: transformer
decoder_conf:
    input_layer: embed
    num_blocks: 6
    linear_units: 2048
    dropout_rate: 0.1
required:
- output_dir
- token_list
version: 0.9.10
distributed: true