Dataset Viewer
Auto-converted to Parquet Duplicate
utt_id
stringlengths
10
10
speaker_id
stringclasses
527 values
sentence
stringlengths
1
125
audio
audioduration (s)
1.1
28.2
__index_level_0__
int64
39
2.53M
f81423c316
8478c
विकास समितिहरूमा र
923,254
4512ed9f33
ece77
मुख्य पृष्ठमा भएको
1,214,837
4b23966741
e72b0
आफू समाजसेवी भएर
645,213
a678173cc6
b5ec7
अभियान पत्रिका एक
324,389
a0573e80fe
55c0f
बहिस्कारको नीति
885,425
859e9a58f4
87f8b
डेङ्गु ज्वरो एक
1,321,105
3c7f1d3077
a30f2
एकमत भएको देखिन्छ
1,288,107
60e7f35612
cbbed
यलम्बरलाई भाषा वंशावलीमा
1,533,308
e0c981aeed
d2aac
विधामा स्नातक
2,345,232
26684e4437
f73cc
प्रतिकार भएको थियो
1,036,586
b787978894
97b71
मेकानिकल कम्पन हावाको
1,215,635
a8001fb549
3240f
जेनिफर लोपेज एक
1,159,683
4bae4c7a0d
1a5c1
नामको पृष्ठ बनाएँ
1,179,494
6ce0e5d345
9c1c2
खेलबाट बाहिरिने छ
1,143,545
127e372c85
7f880
श्रीहरिको कृपाले
2,034,378
4f844ba70a
58fb5
रुकुमको बाँफीकोटमा गरिएको
1,390,219
ab6dbdce53
41b2c
रूसी नष्ट हुन्छन्
2,496,434
c61212a5b6
f07db
भुइँमाथि गुजुल्टो परेको हुन्छ
102,915
ae80674842
d0cd7
खेल खेल्न पाउनेछन्
827,704
9a811f1538
cc4bf
स्थलको प्रयोग गर्नुहोला
152,611
c97495d218
99942
मङ्गल ग्रहले चौथो
577,700
79f842b9cb
2c699
माल्या गर्‍यो
898,169
7cbc660ef9
30a15
विद्यालयबाट स्कुले शिक्षा
766,047
3bcf97eb92
43b70
डन्डीफोर बढ्दाबढ्दै सिस्टको
530,878
89820024ad
58fb5
चिन्नेबास नेपालको गण्डकी अञ्चलको स्याङ्जा जिल्लाको एक गाउँ विकास समिति हो।
204,422
2350d274f0
ef9b4
सफा मौसममा यहाँबाट
773,701
e021501fc7
29c22
उत्साहजनक छ
2,228,861
1a4243b8d8
f45e2
यही भ्रम तोड्न
1,065,112
ff661093f7
bc051
विजय हासिल
1,397,285
d4b94ebf5a
76dab
टेन्चको पुस्तकका डच
1,319,529
c1197d09cf
95a0e
प्रत्येक राष्ट्रको आफ्नै
1,984,552
08d0cbc745
4b140
आयोजना पृष्ठ हुन्
646,471
cec728cd6c
f4b4a
पहिलेदेखि नै धुलिखेलमा
873,119
4ebaf4e148
3867c
धर्मग्रन्थहरूले एउटा मान्छेलाई
860,355
f08c79dfa3
bc6b6
धेरैपछि मात्र पिताजीबाट
545,264
9c3a0de7a3
b66ee
एउटा पात्रमा गहुँ
1,602,510
afdf74b8df
e4714
लेखिएको र साझा
2,499,844
979efc6486
45ef4
एडिलेड सिटी काउन्सिल
1,604,615
9b677392be
f07db
विन्दुमा नयाँ
766,305
3557f3f306
99942
रुकुम जिल्लामा गाभियो
2,021,845
e4bc65e470
07a6a
त्यसपछि उनले त्यस
1,326,635
d3d4e0e10f
22333
आलिङ्गन र चुम्बनका
1,910,037
86abc411cb
4cefa
उच्च माध्यमिक शिक्षा
1,912,306
d0253d1e38
363a2
भारतीय कदमका कारण
2,312,859
beb860e67d
1d9ac
साथै उपत्यकाका अन्य
1,726,064
022fa5675f
43813
अन्तसम्म दक्षिण पूर्वी
954,605
99d35350b6
99866
भारत यात्रामा थिए
733,760
38bdf0121e
bee0c
प्रभुत्व भएको अर्थव्यवस्थामा
248,235
32648ce9a4
b7116
अचानक मुर्छा अवस्थामा
478,336
5416210d45
ea935
नबिनका २ दाजुहरू
546,646
eed029e1f9
39e3b
कुम्भकर्ण एक नेपाली
823,206
31fad8fe2f
83eb0
ठुलोपाखारमा औसत
1,320,157
eb52bc9502
cc4bf
लगाइएको आगो हो
1,076,017
2bbf5d1790
43b70
क्षतिपूर्तिका कतिपय प्रावधानहरू
36,458
2c86898cbc
439c2
नम्बर ५ हो
2,283,798
8ffb641ebb
97b71
नायक भएका छन्
232,438
4abb5b6ac3
b66ee
अन्नको एउटै दाना थिएन।
2,218,003
a14a02c99f
916c6
टाइप गर्न आउँदैन
2,108,570
5ff856836c
ee687
हार्दिक बधाई टक्र्याउँदछु
155,470
b434dc5d30
8cefd
प्रशासनिक जिल्लाहरू मध्येको
564,981
e2da31865e
66260
जुन ४ सम्म गठबन्धन सेनाले ६ सैनिक गुमाएको छ।
1,317,085
de78ea0e9e
6d3f9
माताको नाम राधिका हो
1,844,082
497e835068
83c93
इष्टदेवता रुद्रदेव हुन्
174,330
376c3fa43a
09e5c
यस गाविसमा औसत
429,603
22f33d365d
2f34a
मूल्य प्रसंस्करणमा
899,686
d6a56d4a7b
5f3b4
बैठक काष्ठमण्डप
571,815
792c794488
d31e9
बनाउन पनि सकिन्छ
1,357,873
31a2eba714
1654e
पनि रहेका कारण
1,660,151
f5b8a1b625
72951
गरिदिए सजिलै जान्छन्
986,922
2469b29108
6b883
ध्वस्त भयो
1,177,442
81d8cf79f5
bb466
माघ १ गते
2,317,591
2b45a92114
ccdde
पड्केपछि जसरी नि
1,658,732
2939f93e15
12e6a
ढोका खोलिदियो
2,111,136
c8000ca794
8c66c
पत्ता लगाउन
1,974,755
873ccd8133
44903
उनले स्पष्ट रूपले
1,915,056
df22039061
44903
दिइने सुख सुविधाहरूको
1,014,903
fd65ca27e1
d241a
चिनदेखि सम्पर्क क्षिण
520,118
cffaacc166
53027
परिवर्तित चट्टान हुन्
1,187,548
f1fd5e2bff
64b63
यिनीहरू आफूलाई समाजिक
434,023
ead2343134
aec56
निर्वाचन गर्ने मिति
1,023,538
e0c0b7b983
e7519
जीवनप्रति सकारात्मक धारणाको
2,264,177
c8df2fa5f6
59e9e
ओपन वायर लाइनको
499,894
e09ed7edbc
2ee53
गराएर आफ्ना
670,948
465ff87130
c536c
नेपालीमा एम ए
1,499,177
c77bdc316f
a1143
हिउँदे याममा जाडो छल्न
1,103,301
3051dc7c6b
809f3
जुन थियो नेल्सन
527,671
8f8f4b13c4
25ff3
भाषामा निरुक्तको भूमिकाको
734,142
23fc49a91f
92fd3
जीवित प्राणी लगातार
2,106,485
6d5912ae29
d593e
जेष्ठ तथा असार
1,317,773
e41547c988
1a5c1
काठमाण्डौ स्थित नक्सालको
1,134,736
332541bdf2
73b5e
शक्ति घटेर जान्छ
453,438
0da68c414a
39b91
चिकित्सकहरूको मनमा यो
519,918
89812b03fb
bf591
१९७२ देखि जर्मनीको
2,211,788
1185a999dd
d593e
प्रमुख पनि हुन्थ्यो
722,121
f2b6a44422
5f02c
स्नातक सकेका थिए
2,521,880
170dfd1ad2
41295
बाबुको नाम सूर्य
1,675,398
d9adafdea1
c7d7d
पनि एउटा कच्ची
2,449,001
4edcc03af7
53027
गर्दा म पनि
356,424
34b31f2ea9
3867c
पच्चीस प्रबन्ध
2,289,721
5848eedb10
ca5ed
पानीमा पकाएर बनाइएको
1,663,728
End of preview. Expand in Data Studio

Nepali ASR Dataset (FLAC, Prepared for Whisper Fine-Tuning)

This dataset is a preprocessed and ready-to-use version of the OpenSLR Nepali Automatic Speech Recognition (ASR) corpus, repackaged and standardized to facilitate Whisper model fine-tuning and other speech-to-text experiments.

It provides audio-text pairs in Nepali language, with all audio stored as .flac files and transcriptions in Devanagari script. The dataset follows the Hugging Face datasets format for seamless use with Transformers, SpeechBrain, ESPnet, or fairseq pipelines.


🗂️ Dataset Summary

Split Examples Approx. Size
Train 2,273,832 130.9 GB
Validation 126,324 7.31 GB
Test 126,324 7.24 GB
Total ~2.52M ~145.48 GB

All audio files have been resampled or confirmed to use a 16 kHz sampling rate to align with Whisper’s input requirements.


🧾 Data Fields

Column Type Description
utt_id string Unique utterance ID
speaker_id string Speaker identifier
sentence string Transcribed Nepali sentence
audio Audio (16 kHz) FLAC audio waveform
__index_level_0__ int64 Internal dataset index

📦 Dataset Structure

After extraction, the dataset directory was organized as:

extracted/
 ├── asr_nepali_0/
 │   └── asr_nepali/data/
 │       ├── 00/
 │       │    ├── *.flac
 │       ├── 01/
 │       ├── ...
 │       ├── 0f/
 ├── asr_nepali_1/
 │   └── asr_nepali/data/
 │       ├── 00/ ... 0f/
 ...

A master metadata file (master.tsv) was programmatically generated to map all audio files and their corresponding transcripts, later split into train/validation/test subsets.


🧠 Intended Use

  • Fine-tuning Whisper, Wav2Vec2, or Conformer models for Nepali ASR tasks.
  • Benchmarking speech recognition and language modeling in Nepali.
  • Supporting RAG (Retrieval-Augmented Generation) speech-text pipelines or multilingual model adaptation.

⚙️ Example Usage

from datasets import load_dataset

dataset = load_dataset("Firoj112/nepali-asr-flac", split="train")
print(dataset[0])
# {
#   'utt_id': 'utt_001',
#   'speaker_id': 'spk_12',
#   'sentence': 'नेपाल सुन्दर देश हो ।',
#   'audio': {'array': [...], 'sampling_rate': 16000}
# }

🧩 Technical Notes

  • All .flac files were loaded using soundfile (sf.read) to ensure consistent audio arrays.
  • Dataset split ratio: Train 90% / Validation 5% / Test 5%
  • Audio and transcription alignment verified through automatic parsing of OpenSLR transcripts.

📚 Source

This dataset is based on the OpenSLR Nepali ASR corpus, distributed under the Creative Commons Attribution-ShareAlike 4.0 International License.

Source: https://www.openslr.org/54


⚖️ License

Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0)

This dataset inherits its license from the original OpenSLR Nepali ASR release. You are free to share and adapt the material for any purpose, even commercially, under the following terms:

  • Attribution — You must give appropriate credit to the original OpenSLR dataset.
  • ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license.

See the full license text here.


🧑‍💻 Acknowledgements

  • OpenSLR contributors for curating and publishing the original Nepali ASR corpus.
  • This repackaged dataset was created to simplify Whisper and multilingual ASR fine-tuning workflows.

🚀 Citation

If you use this dataset, please cite both OpenSLR and this preprocessed release.

@misc{openslr_nepali_asr,
  title = {Nepali ASR Corpus (OpenSLR #114)},
  author = {OpenSLR Contributors},
  year = {2022},
  url = {https://www.openslr.org/114},
  note = {Licensed under CC BY-SA 4.0}
}

@misc{nepali_asr_flac_prep,
  title = {Nepali ASR (FLAC) - Whisper Fine-Tuning Ready Dataset},
  author = {SAGEA},
  year = {2025},
  note = {Processed and released on Hugging Face Datasets}
}

Summary

  • 🌐 Language: Nepali (ne)
  • 🔊 Format: .flac (16 kHz)
  • 🧾 License: CC BY-SA 4.0
  • 🧠 Task: Automatic Speech Recognition (ASR)
  • ⚡ Ready for: Whisper, Wav2Vec2, Conformer, SpeechT5

Prepared and cleaned for research and educational purposes.

Downloads last month
317