Datasets:

sulabhkatiyar
/

ne-asr-dataset-nre-aug

text stringlengths 29 254	language stringclasses 1 value	augmentation stringclasses 3 values
asika water tank zu pe fill thyu kenjün tsü hyu bin	Rengma	original
asika water tank zu pe fill thyu kenjün tsü hyu bin	Rengma	speed_0.9
asika water tank zu pe fill thyu kenjün tsü hyu bin	Rengma	speed_1.1
side nme wange ne teru dressing thyu thyu hon machine ne binyon	Rengma	original
side nme wange ne teru dressing thyu thyu hon machine ne binyon	Rengma	speed_0.9
side nme wange ne teru dressing thyu thyu hon machine ne binyon	Rengma	speed_1.1
tsü ka tsü a kejwen npugü le dun kebin tsü hu	Rengma	original
tsü ka tsü a kejwen npugü le dun kebin tsü hu	Rengma	speed_0.9
tsü ka tsü a kejwen npugü le dun kebin tsü hu	Rengma	speed_1.1
lo side ki tsü hmm shunbin ne bin bin nyon lo wire den tsü ku bin bin nyon	Rengma	original
lo side ki tsü hmm shunbin ne bin bin nyon lo wire den tsü ku bin bin nyon	Rengma	speed_0.9
lo side ki tsü hmm shunbin ne bin bin nyon lo wire den tsü ku bin bin nyon	Rengma	speed_1.1
temi nme tsüle dun o kedun ben nyu a phone nnri o khwen bin	Rengma	original
temi nme tsüle dun o kedun ben nyu a phone nnri o khwen bin	Rengma	speed_0.9
temi nme tsüle dun o kedun ben nyu a phone nnri o khwen bin	Rengma	speed_1.1
tin ne ahh hyü bin no tokta nme le kepha bin	Rengma	original
tin ne ahh hyü bin no tokta nme le kepha bin	Rengma	speed_0.9
tin ne ahh hyü bin no tokta nme le kepha bin	Rengma	speed_1.1
no hhmm asheki shunbin den le kethon kethon no kechon le bin	Rengma	original
no hhmm asheki shunbin den le kethon kethon no kechon le bin	Rengma	speed_0.9
no hhmm asheki shunbin den le kethon kethon no kechon le bin	Rengma	speed_1.1
ashyeki jentin che kebin tsü ku hyu	Rengma	original
ashyeki jentin che kebin tsü ku hyu	Rengma	speed_0.9
ashyeki jentin che kebin tsü ku hyu	Rengma	speed_1.1
ki teshinki aah khujal le martol le rod le cho kan kechacha jwen nyon	Rengma	original
ki teshinki aah khujal le martol le rod le cho kan kechacha jwen nyon	Rengma	speed_0.9
ki teshinki aah khujal le martol le rod le cho kan kechacha jwen nyon	Rengma	speed_1.1
tsü ase ki hhmm günyo bin no thai bin no ka colour ken kenhen cha kebin hyu	Rengma	original
tsü ase ki hhmm günyo bin no thai bin no ka colour ken kenhen cha kebin hyu	Rengma	speed_0.9
tsü ase ki hhmm günyo bin no thai bin no ka colour ken kenhen cha kebin hyu	Rengma	speed_1.1
pechenyu nme le shu binyo lesu tachepen o	Rengma	original
pechenyu nme le shu binyo lesu tachepen o	Rengma	speed_0.9
pechenyu nme le shu binyo lesu tachepen o	Rengma	speed_1.1
box nyuki le light light ron den ne bin	Rengma	original
box nyuki le light light ron den ne bin	Rengma	speed_0.9
box nyuki le light light ron den ne bin	Rengma	speed_1.1
pe anyu la tha pe anyu ben tsü lalu anyu ne paha le apha penshu o tesoi tsü n nri bin	Rengma	original
pe anyu la tha pe anyu ben tsü lalu anyu ne paha le apha penshu o tesoi tsü n nri bin	Rengma	speed_0.9
pe anyu la tha pe anyu ben tsü lalu anyu ne paha le apha penshu o tesoi tsü n nri bin	Rengma	speed_1.1
amen ka hi nyu ne hindu nden kereshvu ben tsü hyu	Rengma	original
amen ka hi nyu ne hindu nden kereshvu ben tsü hyu	Rengma	speed_0.9
amen ka hi nyu ne hindu nden kereshvu ben tsü hyu	Rengma	speed_1.1
atsü anyun hon instruments nden kechacha pe ben kenjün	Rengma	original
atsü anyun hon instruments nden kechacha pe ben kenjün	Rengma	speed_0.9
atsü anyun hon instruments nden kechacha pe ben kenjün	Rengma	speed_1.1
next picture ka hi nri keshü busy market nme hon che kethyu	Rengma	original
next picture ka hi nri keshü busy market nme hon che kethyu	Rengma	speed_0.9
next picture ka hi nri keshü busy market nme hon che kethyu	Rengma	speed_1.1
ro aka le hmm building kenshen penshu binyon	Rengma	original
ro aka le hmm building kenshen penshu binyon	Rengma	speed_0.9
ro aka le hmm building kenshen penshu binyon	Rengma	speed_1.1
nchenyu kan jwen nyon abola da binyon anya da binyon	Rengma	original
nchenyu kan jwen nyon abola da binyon anya da binyon	Rengma	speed_0.9
nchenyu kan jwen nyon abola da binyon anya da binyon	Rengma	speed_1.1
aka hi picture ka hi aahh tem mele chair ka dun nyo pechenyu ne dun nyo tenunyu da	Rengma	original
aka hi picture ka hi aahh tem mele chair ka dun nyo pechenyu ne dun nyo tenunyu da	Rengma	speed_0.9
aka hi picture ka hi aahh tem mele chair ka dun nyo pechenyu ne dun nyo tenunyu da	Rengma	speed_1.1
tsü nyeki kam ki mache n chanyu mehvu le bin kebin tsü hyu	Rengma	original
tsü nyeki kam ki mache n chanyu mehvu le bin kebin tsü hyu	Rengma	speed_0.9
tsü nyeki kam ki mache n chanyu mehvu le bin kebin tsü hyu	Rengma	speed_1.1
ro tsü side nmeo ahon opposite wange gari white colour nmele parking thyupe ben binyon	Rengma	original
ro tsü side nmeo ahon opposite wange gari white colour nmele parking thyupe ben binyon	Rengma	speed_0.9
ro tsü side nmeo ahon opposite wange gari white colour nmele parking thyupe ben binyon	Rengma	speed_1.1
weather che meche foggy ketsin choo aahh nyen pe ben kebin tsü picture hika bin	Rengma	original
weather che meche foggy ketsin choo aahh nyen pe ben kebin tsü picture hika bin	Rengma	speed_0.9
weather che meche foggy ketsin choo aahh nyen pe ben kebin tsü picture hika bin	Rengma	speed_1.1
aka amen ka n nyu ne hyu kebin ne temi aden thyü o kekwen kebin tsü hyu bin	Rengma	original
aka amen ka n nyu ne hyu kebin ne temi aden thyü o kekwen kebin tsü hyu bin	Rengma	speed_0.9
aka amen ka n nyu ne hyu kebin ne temi aden thyü o kekwen kebin tsü hyu bin	Rengma	speed_1.1
cho lesu nden nmvu ai zu nden necessities den stationaries nden kejwen ben nme honche kethyu tsü hyu ro	Rengma	original
cho lesu nden nmvu ai zu nden necessities den stationaries nden kejwen ben nme honche kethyu tsü hyu ro	Rengma	speed_0.9
cho lesu nden nmvu ai zu nden necessities den stationaries nden kejwen ben nme honche kethyu tsü hyu ro	Rengma	speed_1.1
teso ki tsü ka den tsü ku building den tsü ku binyon	Rengma	original
teso ki tsü ka den tsü ku building den tsü ku binyon	Rengma	speed_0.9
teso ki tsü ka den tsü ku building den tsü ku binyon	Rengma	speed_1.1
picture aka hi aro hmm tin ne binyon	Rengma	original
picture aka hi aro hmm tin ne binyon	Rengma	speed_0.9
picture aka hi aro hmm tin ne binyon	Rengma	speed_1.1
acho kechon ne binyon ahh acho paha le wall thyu shü binyon paha le acho kedada le binyon	Rengma	original
acho kechon ne binyon ahh acho paha le wall thyu shü binyon paha le acho kedada le binyon	Rengma	speed_0.9
acho kechon ne binyon ahh acho paha le wall thyu shü binyon paha le acho kedada le binyon	Rengma	speed_1.1
tenunyu nme le binyon hmm akan white colour le pink colour a u pvu nyon	Rengma	original
tenunyu nme le binyon hmm akan white colour le pink colour a u pvu nyon	Rengma	speed_0.9
tenunyu nme le binyon hmm akan white colour le pink colour a u pvu nyon	Rengma	speed_1.1
tsü nyeki asika gun gü bench so nyu tsü seat cushion hyu	Rengma	original
tsü nyeki asika gun gü bench so nyu tsü seat cushion hyu	Rengma	speed_0.9
tsü nyeki asika gun gü bench so nyu tsü seat cushion hyu	Rengma	speed_1.1
senkethyu nyu den tsü le soo photo lo kebin mehvu le photo lo kebin tsü hyu	Rengma	original
senkethyu nyu den tsü le soo photo lo kebin mehvu le photo lo kebin tsü hyu	Rengma	speed_0.9
senkethyu nyu den tsü le soo photo lo kebin mehvu le photo lo kebin tsü hyu	Rengma	speed_1.1
hhmm nchenyu den le track pant pvü	Rengma	original
hhmm nchenyu den le track pant pvü	Rengma	speed_0.9
hhmm nchenyu den le track pant pvü	Rengma	speed_1.1
ro tsü hensin kitsü pechenyu den tenunyu den ku binyon paha le azü khipe elephant hyu binyon	Rengma	original
ro tsü hensin kitsü pechenyu den tenunyu den ku binyon paha le azü khipe elephant hyu binyon	Rengma	speed_0.9
ro tsü hensin kitsü pechenyu den tenunyu den ku binyon paha le azü khipe elephant hyu binyon	Rengma	speed_1.1
lo side kistü temi kechon ne dun binyon	Rengma	original
lo side kistü temi kechon ne dun binyon	Rengma	speed_0.9
lo side kistü temi kechon ne dun binyon	Rengma	speed_1.1
chenbi ahh tha nme anye anye ku bin ro shunbin den ku binyon	Rengma	original

End of preview. Expand in Data Studio

NE ASR Augmented Dataset -- Rengma (nre)

Augmented automatic speech recognition dataset for Rengma (nre), a Tibeto-Burman language spoken in Nagaland, India.

Source

Augmented from sulabhkatiyar/ne-asr-nre (original transcribed speech data from the ARTPARK-IISc Vaani project).

Language Information

Property	Value
Language	Rengma
ISO 639-3	`nre`
Family	Tibeto-Burman
Region	Nagaland, India
Tonal	Yes
Tier	B (0.51h original data)

Dataset Statistics

Original training samples: 336
Augmented training samples: 1,008 (3x augmentation)
Train shards: 2
Estimated original duration: ~0.5 hours
Estimated augmented duration: ~1.5 hours

Split	Samples
train	1,008
validation	11
test	28

Transformations Applied

Each original training sample produces 3 samples (1 original + 2 speed + 0 pitch):

Speed perturbation: 0.9x, 1.1x (2 variants per sample)
Pitch shift: Disabled (tonal language -- pitch shift would alter lexical meaning)
Noise augmentation: Not applied

SpecAugment Parameters (for training, NOT in this dataset)

These parameters are consumed by the training script and are not baked into the audio files:

mask_time_prob: 0.05
mask_time_length: 10
mask_feature_prob: 0.03
mask_feature_length: 10
layerdrop: 0.05

Full augmentation config: configs/augmentation_config.yaml

Dataset Format

Audio: 16kHz mono WAV (stored as Parquet with audio bytes)
Text: Transcriptions
Features: audio, text, language, augmentation
Augmentation labels: original, speed_0.9, speed_1.1

How to Use

from datasets import load_dataset

# Load the full dataset
ds = load_dataset("sulabhkatiyar/ne-asr-nre-aug")

# Load only the training split
train = load_dataset("sulabhkatiyar/ne-asr-nre-aug", split="train")

# Filter to only original (non-augmented) samples
original_only = train.filter(lambda x: x["augmentation"] == "original")

# Filter to a specific augmentation type
speed_09 = train.filter(lambda x: x["augmentation"] == "speed_0.9")

Original Data

Source dataset: sulabhkatiyar/ne-asr-nre
Project: ARTPARK-IISc Vaani
License: CC-BY-4.0

Citation

If you use this dataset, please cite the Vaani project and acknowledge the augmentation pipeline.

Downloads last month: 40

Models trained or fine-tuned on sulabhkatiyar/ne-asr-dataset-nre-aug

sulabhkatiyar/ne-asr-nre

Automatic Speech Recognition • Updated 4 days ago • 32