Hugging Face's logo

Datasets:
bookbot
/
slr72-es-ipa-v1

Tasks:
Automatic Speech Recognition
Modalities:
Audio
Text
Formats:
parquet
Languages:
Spanish
Size:
1K - 10K
Tags:
speech
phonemes
ipa
Libraries:
Datasets
Dask
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
Dataset Viewer
Auto-converted to Parquet Duplicate
id
stringlengths
23
26
audio
audioduration (s)
1.96
13.8
ipa_transcript
stringlengths
19
249
com_07508_01944220782-3612
k j e ɾ o i ɾ m e d e b a k a s j o n e s a a w a j p e ɾ o t e n g o e s t a d u d a d e s i e l b o l k a n e s t a a k t i b o
cof_09697_00540469455-2024
e n l a s e g u n d a f i l a k e d a n t ɾ e s e l u g a ɾ e s l i b ɾ e s
com_06136_00103025283-703
e n k e o t ɾ a s p e l i k u l a s s a l e i n t e ɾ p ɾ e t a d o e l p e ɾ s o n a x e d e l e n i n
cof_03034_01667337711-3377
m e g u s t a t o m a ɾ a g w a m i n e ɾ a l
com_09334_00347372707-671
l a s t e x a s r o x a s d e l t e x a d o n o s o n d e m e x i k o s o n d e t e k s a s
com_02121_00344397159-1571
l a n a b e e n k a ʝ o e n l o s a r e s i f e s
com_06136_01336462270-3681
k w a n t o k w e s t a u n a b i s i k l e t a u s a d a
com_07049_00137018832-672
l a s j u d a d e s b e ʝ i s i m a i t a n a l e g ɾ e
cof_09697_01349363078-25
e n k e k a ʝ e e s t a l a k a s a
com_09697_00809791462-881
r e m e d j o s b a ɾ o f i n a l i s o e l m u ɾ a l e n l o n d ɾ e s e n m i l n o b e s j e n t o s s i n k w e n t a i o t͡ʃ o
com_09334_00033616669-3024
a j a l g u n a o b ɾ a d e t e a t ɾ o e s t e f i n d e s e m a n a
cof_01523_00953609331-355
l a p e l i k u l a m e ʝ e g o a l k o ɾ a s o n
com_07049_01438121780-1591
e s r e k o m e n d a b l e t o m a ɾ t e k a l j e n t e s i n k a f e i n a a n t e s d e d o ɾ m i ɾ
cof_09697_00912368462-3118
ʝ a n o k j e ɾ o u s a ɾ t a n t o m i k o t͡ʃ e k j e ɾ o u s a ɾ m a s l a b i s i k l e t a
com_02484_00834333475-295
m e p a ɾ e s e m w i i n t e ɾ e s a n t e l a i m f o ɾ m a s j o n
cof_05223_00824464203-1470
a j a l g u n a o p s j o n e n e l m e n u p a ɾ a k j e n n o k o m e k a ɾ n e
cof_09334_01532357372-3760
m e p w e d e s d e s i ɾ k e t e l e f o n o d e b o k o m p ɾ a ɾ
com_03397_00285803202-225
e l f u t b o l t͡ʃ i n o a e b o l u s j o n a d o a n i b e l e s n u n k a a n t e s b i s t o s
com_02436_02006692536-1866
k w a l e s s o n l a s n o t i s j a s m a s r e s j e n t e s d e l f w e g o f o ɾ e s t a l
cof_03397_01045413276-3249
t e n g o u n a r e w n j o n o j e n m i k a s a p e ɾ o n o s e k e d a ɾ d e t o m a ɾ
cof_07049_01837393219-2770
t e b o j a m a n d a ɾ e l m e n u d e l a s e n a a t u k o r e o e l e k t ɾ o n i k o
cof_09334_00249065004-3816
u l t i m a m e n t e e b i s t o m u t͡ʃ a s b i s i k l e t a s r o x a s
com_01523_00461019966-1209
p e ɾ d o n k e t e i n t e r u m p a p e ɾ o n o s e k e d i a e s
com_07049_01312042990-1540
l a n a b e e n k a ʝ o e n l o s a r e s i f e s
cof_08421_01522933317-2854
k j e n d e l a l i s t a d a e l m e x o ɾ m a s a x e f a s j a l
com_09697_01522972535-2535
a j a l g u n b w e l o m a s e k o n o m i k o p a s a n d o p o ɾ l a s i s l a s k a n a ɾ j a s
cof_03349_01402773241-333
t o d a l a k o m i d a t͡ʃ a t a r a e s f a l s a e s t o t a l m e n t e k i m i k a
com_06136_01252022054-1015
k o m o p w e d o r e x i s t ɾ a ɾ m e p a ɾ a e n t ɾ a ɾ a l a u n i b e ɾ s i d a d
cof_07508_01657429402-3382
e s t o j p o ɾ m u d a ɾ m e i k j e ɾ o b u s k a ɾ u n n w e b o d e p a ɾ t a m e n t o d o n d e a s e p t e n m a s k o t a s
com_03349_02054918200-2945
u l t i m a m e n t e e b i s t o m w i p o k a s b i s i k l e t a s e n l a s j u d a d
com_00610_01550783796-986
a k a e n t ɾ e n o s e s t a b w e n a l a p i s s a
com_03349_00455164343-1041
k w a n t o k w e s t a e s e t ɾ a x e d e b a ɲ o e s p a ɲ o l
cof_01523_01633943376-3250
k w a n d o r e g ɾ e s a e l p ɾ e s e n t a d o ɾ d e n o t i s j a s d e s u s b a k a s j o n e s
com_07049_00533940188-968
k e r e s t a w ɾ a n t e e s e l k e m a s m e r e k o m j e n d a s
cof_02436_00262476365-2700
s i t e m a ɾ e a s f a s i l e n u n b a ɾ k o p w e d e s ʝ e g a ɾ p o ɾ t j e r a ʝ a k e n o e s u n a i s l a s i n o u n a p e n i n s u l a
cof_09334_00842347579-1515
k j e ɾ e s k e t e s i ɾ b a u n k a f e
com_03349_00503313664-1884
l a t j e n d a d e x a b o n e s i p e ɾ f u m e s f i n o s t j e n e r e g a l o s p a ɾ a e l d i a d e l a s m a d ɾ e s
cof_04310_00813070265-60
s o s t e n l a p a t a d e l a s i ʝ a k o n t u m a n o d e ɾ e t͡ʃ a i t o m a e l b o t e d e p e g a m e n t o k o n t u m a n o i s k j e ɾ d a
com_04310_01585798896-2069
k j e ɾ o m i a s j e n t o e n e l p a s i ʝ o
cof_05223_01879611762-2446
l a g ɾ a n i s a d a d e s t ɾ u ʝ o t o d a l a p l a n t a s j o n d e l e t͡ʃ u g a
cof_09697_02107402353-2739
e l p u b l i k o e s t a e n l o k e s i d o
com_07508_01301656635-3239
n o s u p o d e s s i f ɾ a ɾ l a s w e ʝ a s
com_09697_00404046635-165
l e r e k o m j e n d o l o s f j o ɾ d o s n o ɾ w e g o s
com_06136_01783429805-3712
d o n d e e s t a n l a s e s t a s j o n e s d e b i s i k l e t a s d e l a s j u d a d
com_03349_01303717992-523
n o e n t j e n d o l a d i f e ɾ e n s j a e n t ɾ e l o s d i f e ɾ e n t e s x e n e ɾ o s t e a t ɾ a l e s
cof_06136_00931710764-2658
k w a n t o k w e s t a e l e s t a s j o n a m j e n t o p o ɾ o ɾ a
com_07049_00213617479-2743
k j e ɾ o a s e ɾ a l g o m e n o s i n t e n s o
cof_00610_01936293184-528
a o ɾ a r e b i s o l a i m f o ɾ m a s j o n k e m e m a n d a s t e i l o r e s w e l b o
cof_04310_01201344549-1413
e s t o j b j e n e m o s j o n a d a m e b o j a m a r w e k o s
com_00610_00269776593-522
d i m e a s j e n s j a s j e ɾ t a k j e n e s e s t e p e ɾ s o n a x e
cof_09697_00433496049-3139
e s t o j t e n j e n d o p ɾ o b l e m a s k o n m i k o t͡ʃ e i u n a m i g o m e r e k o m e n d o u n f e r a ɾ i
com_04310_01707318563-1298
e l x u g a d o ɾ s a l j o e n e l m i n u t o u n o p o ɾ u n a m o l e s t j a e n l a k a b e s a
cof_05223_01537762584-3273
l o s k o ɾ a s o n e s d e p o ʝ o s o n u n a d e l i s j a
com_03349_00575481758-1938
m e p a ɾ e s e m w i b j e n
cof_03034_00517500155-2711
k w a n d o n a s j o n a p o l e o n b o n a p a ɾ t e
com_03034_00359508017-3554
p a ɾ a a b ɾ i ɾ l o s p o ɾ o s d e l a p j e l l o m e x o ɾ e s i ɾ a u n s a w n a a u n b a p o ɾ
cof_03349_02103001645-3127
n o s e s i t e b a ʝ a a b e ɾ m a ɲ a n a
cof_03034_01070846555-2653
l a p ɾ o n u n s j a s j o n d e l f ɾ a n s e s e s d i f i s i l
com_04310_01249858893-2769
l a s b i s a g ɾ a s s o n a s u l e s
cof_03034_00503729338-2114
ʝ o n a s i e s e m i s m o a ɲ o
com_02484_00378640484-2237
p ɾ i m e ɾ o k e n a d a b a m o s a a s e ɾ u n e x e ɾ s i s j o d e r e s p i ɾ a s j o n
cof_00610_00749562938-2168
t e r e k o m j e n d o k e u s e s p a n t a l o n e s m j e n t ɾ a s t e n g a s e s a e ɾ i d a e n l a r o d i ʝ a
com_09697_02000930211-3657
p o ɾ l a s t a ɾ d e s s e e s k u t͡ʃ a m u s i k a e n l a p l a s a
cof_05223_00804564149-2160
m e a k a b o d e k o m p ɾ a ɾ u n r o m p e k a b e s a s
com_02436_00958421787-678
l a k a ʝ e e s t a ʝ e n a d e o ʝ o s
com_05223_01237219710-1613
k w a l e s e l o s p i t a l m a s s e ɾ k a n o
com_04310_01075339364-68
l o k e n o m b ɾ a s t e s o n p e l i k u l a s o s o n s e ɾ j e s
com_03034_00323142621-177
k e e s u n a t a x o
cof_09334_01667729988-2367
e l ʝ a n o e n ʝ a m a s e s u n k l a s i k o d e x w a n r u l f o
cof_08784_01727477629-2192
e l f ɾ a n s e s e s e l i d j o m a d e l a m o ɾ
com_09697_00429488546-3600
m a ɲ a n a b a a a s e ɾ s o l p o ɾ l a m a ɲ a n a
cof_09334_01762674784-1671
k w e s t a s j e t e m i l p e s o s p e ɾ o n o k w e s t a e l e m b i o
com_02484_00233158256-882
m e p w e d e s r e p e t i ɾ l a i m f o ɾ m a s j o n p o ɾ f a b o ɾ
cof_07049_02003227126-1126
a l p a ɾ e s e ɾ t ɾ u m p i l a a l k a l d e s a d e p w e ɾ t o r i k o s e e m f ɾ e n t a ɾ o n e n t w i t t e ɾ
com_04310_00544305790-2352
k e o p s j o n e s d e o ɾ a ɾ j o d e s a l i d a m e o f ɾ e s e s
com_09334_01703438469-3537
t o k a ɾ e l k s i l o f o n o e s m i o b b i f a b o ɾ i t o
com_03349_00030288521-2429
l a k a ʝ e e s t a ʝ e n a d e o ʝ o s
cof_07508_00942130171-680
n o t e n g o n i i d e a d e k o m o k a m b j a ɾ u n a r w e d a n o s e s i p w e d a s p e d i ɾ a s i s t e n s j a o m e p w e d e s a ʝ u d a ɾ p o ɾ f a
com_03034_01046005923-2523
a s m e l a s i t a k o n e l m a e s t ɾ o d e x w a n
cof_01523_01157644137-3336
e n l a s n o t i s j a s d e l a n o t͡ʃ e n o b i a l p ɾ e s e n t a d o ɾ d e s j e m p ɾ e
com_05223_01588380116-2485
n o t e e n t j e n d o n a d a
cof_01523_02053596870-2360
e n e s t e m o m e n t o e s t o j a s j e n d o t u k o m p ɾ a
com_03397_01972609978-1154
p a ɾ a a s e ɾ e s a r e s e t a n e s e s i t a s k a ɾ n e d e k a l i d a d s u p e ɾ j o ɾ
cof_08421_00644498008-1956
l a r e p o s t e ɾ i a e s u n a ɾ t e
cof_02436_01862493241-1528
k w a n t o s d i a s k j e ɾ e s e s t a ɾ f w e ɾ a
com_03397_01799593922-1589
o x a l a p u d j e ɾ a m o s i m b i t a ɾ a e s t a x u n t a k o n l a s e ɲ o ɾ a a ɾ j a s a b a ɾ j o s p i n t o ɾ e s
cof_04310_00556000594-32
l a k a ʝ e e s t a ʝ e n a d e o ʝ o s
cof_03034_00475056748-317
k w a l e s e l k o l o ɾ f a b o ɾ i t o d e l a s n i ɲ a s
com_04310_01686162666-292
t o d a e s t a i m f o ɾ m a s j o n l a p w e d e e n k o n t ɾ a ɾ e n l a p a x i n a w e b d e l a s e k ɾ e t a ɾ i a d e e d u k a s j o n
cof_09334_01387063792-1018
k j e n s u b j o e l m e x o ɾ b i d e o
com_03397_02021731341-412
s e g u n m i s d a t o s e s t e m j e ɾ k o l e s o t͡ʃ o d e m a ʝ o n o a j p a ɾ t i d o s d e t e n i s
cof_09334_01419340408-1183
k j e ɾ o k e l a b o l s a s e p w e d a u s a ɾ d e p a ɲ a l e ɾ a p e ɾ o k e t a m b j e n l a p w e d a u s a ɾ p a ɾ a s a l i ɾ
com_03397_01806966927-1945
p a ɾ a a s e ɾ e s a r e s e t a n e s e s i t a s a s e j t e d e o l i b a i a s e j t e d e x i ɾ a s o l
com_00610_00113450014-233
m e a k a b o d e k o m p ɾ a ɾ u n p ɾ o ʝ e k t o ɾ d e d j a p o s i t i b a s p e ɾ o n o t e n g o i d e a d e k o m o a ɾ m a ɾ l o
com_07508_00373090508-2796
t u m e p o d ɾ i a s d a ɾ a l g u n o s n o m b ɾ e s d e r a s a s p e k e ɲ a s
com_09697_01632243100-781
e l p i x a m a d e r a ʝ a s e s a s u l
com_05223_01773075287-2650
ʝ a b i l a s f o t o s k e s u b i s t e a ʝ e ɾ a l a r e d t e b e s m w i b j e n t e k e d a m w i b j e n e l k o l o ɾ r o x o
com_08784_01981551433-2759
t j e n e n k e e s t a ɾ d o s o ɾ a s a n t e s d e l a f u n s j o n p a ɾ a r e k o x e ɾ l o s b o l e t o s
com_07508_01163467443-951
e s k e s o ɲ e k o n e s t e a k t o ɾ e s t e a k t o ɾ k e s a l e e n l a p e l i k u l a e s t a d e l p e t ɾ o l e o s a n g ɾ j e n t o
cof_09697_02119101828-760
e l k a b a ʝ o e s t a a m a r a d o
End of preview. Expand in Data Studio

OpenSLR72 Colombian Spanish — canonical IPA v1

This repository is the approved deterministic v1 publication cut of the OpenSLR72 Colombian Spanish recipe data. It contains the accepted final Lhotse cuts only:

split rows
train 3,922
test 981

The data files have exactly three columns:

  • id — the unique final Lhotse cut identifier (for example, com_04310_01996434277-0).
  • audio — a Hugging Face Audio feature containing the original encoded source bytes and source basename.
  • ipa_transcript — the exact whitespace-separated canonical v1 phone string from supervisions[0].text in the final cut.

The legacy source Parquet field phonemes_ipa (BabyGruut-era output) was not used. The IPA strings here are the v1 canonical recipe supervision, validated against the fixed atomic phone inventory in egs/bookbot_es/ASR/es-tokens.txt.

Source, license, attribution, and citation

The source is Crowdsourced high-quality Colombian Spanish speech data set (OpenSLR SLR72), originally collected by Google and distributed through OpenSLR. The source license is Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0). The pinned source manifest is slr72_dataset/manifest.tsv, SHA-256 fda09e2a168617d6eebabe6553b93ee0b6cce654d26e050cb3482b41718c7d27; the pinned source revision is e7c9924b964eecbdcba1181add860367d3c7ca3a.

Please preserve this attribution, identify the v1 IPA transformation described below, and distribute adaptations under CC BY-SA 4.0. The source page records the following citation:

@inproceedings{guevara-rukoz-etal-2020-crowdsourcing,
  title = {{Crowdsourcing Latin American Spanish for Low-Resource Text-to-Speech}},
  author = {Guevara-Rukoz, Adriana and Demirsahin, Isin and He, Fei and Chu, Shan-Hui Cathy and Sarin, Supheakmungkol and Pipatsrisawat, Knot and Gutkin, Alexander and Butryna, Alena and Kjartansson, Oddur},
  booktitle = {Proceedings of The 12th Language Resources and Evaluation Conference (LREC)},
  year = {2020},
  month = may,
  address = {Marseille, France},
  publisher = {European Language Resources Association (ELRA)},
  url = {https://www.aclweb.org/anthology/2020.lrec-1.801},
  pages = {6504--6513},
  ISBN = {979-10-95546-34-4},
}

Preparation and transformations

  • Source rows were joined by the pinned audio.path basename to the final cut supervision/recording identity.
  • Every accepted cut has start = 0 and covers its complete source recording. Therefore the embedded encoded source bytes are retained losslessly; there were zero trimmed cuts requiring extraction.
  • id is the final cut ID, not a regenerated source ID.
  • ipa_transcript is copied exactly from the final canonical supervision and is not reconstructed from orthographic text or the legacy phonemes_ipa field.
  • The final cut metadata records the training recipe’s optional resample/volume/speed transforms, but those augmentations are intentionally not applied to this source-preserving publication.

The source manifest’s Charsiu raw phonemization is provenance only; downstream v1 normalization and rejection decisions are those recorded in the final cuts. The exact output shard hashes, row counts, cut manifest hashes, source revision, and transformation policy are machine-readable in provenance/provenance.json; the complete one-to-one source join audit is in provenance/join_manifest.tsv.gz.

Fixed smoke sample

The Common Voice smoke sample is not part of this repository. Runtime tests for this SLR72 package should decode a copied waveform at 16 kHz; this dataset preserves each original 48 kHz WAV source byte stream.

License notice

This derived dataset remains under CC BY-SA 4.0 because it contains the source audio/transcriptions and a canonical IPA adaptation. See LICENSE and the OpenSLR72 source page for the original notices.
Downloads last month
77
Number of rows:
4,903
Total file size:
1.62 GB

Models trained or fine-tuned on bookbot/slr72-es-ipa-v1