Datasets:

sdmy
/

kokborok

Modalities:

Text

ArXiv:

Dataset card Data Studio Files Files and versions Community

Dataset Preview

Split (1)

train

The full dataset viewer is not available (click to read why). Only showing a preview of the rows.

The dataset generation failed because of a cast error

Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 4 new columns ({'is_src_orig', 'id', 'factuality', 'srcs'}) and 2 missing columns ({'is_source_orig', 'src'}).

This happened while the json dataset builder was generating data using

hf://datasets/sdmy/kokborok/smoldoc__en_trp.json (at revision d681e40a913f1c559156e2b428e322ec5abc0308)

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1870, in _prepare_split_single
                  writer.write_table(table)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 622, in write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2292, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2240, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              id: string
              sl: string
              tl: string
              srcs: list<item: string>
                child 0, item: string
              trgs: list<item: string>
                child 0, item: string
              factuality: string
              is_src_orig: bool
              -- schema metadata --
              pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 897
              to
              {'src': Value(dtype='string', id=None), 'trgs': Sequence(feature=Value(dtype='string', id=None), length=-1, id=None), 'sl': Value(dtype='string', id=None), 'tl': Value(dtype='string', id=None), 'is_source_orig': Value(dtype='bool', id=None)}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1438, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1050, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 924, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1000, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1741, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1872, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 4 new columns ({'is_src_orig', 'id', 'factuality', 'srcs'}) and 2 missing columns ({'is_source_orig', 'src'}).
              
              This happened while the json dataset builder was generating data using
              
              hf://datasets/sdmy/kokborok/smoldoc__en_trp.json (at revision d681e40a913f1c559156e2b428e322ec5abc0308)
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

src string	trgs sequence	sl string	tl string	is_source_orig bool
a	[ "kaisa", "masa", "Khoroksa" ]	en	trp	true
a little	[ "kisasa" ]	en	trp	true
a lot	[ "kwbangma" ]	en	trp	true
a lot of	[ "kwbangma" ]	en	trp	true
abandon	[ "yakarwi thangmani" ]	en	trp	true
abandoned	[ "khibi kwlangjak" ]	en	trp	true
abide	[ "kok khwnai tongmani" ]	en	trp	true
ability	[ "chukmung" ]	en	trp	true
able	[ "chukjak" ]	en	trp	true
about	[ "no twiwi" ]	en	trp	true
above	[ "sakao" ]	en	trp	true
abroad	[ "phatar hathai" ]	en	trp	true
absence	[ "manjagwiya", "kwrwi wngmani" ]	en	trp	true
absent	[ "manjakphaiya", "kwrwi" ]	en	trp	true
absolutely	[ "Ahaino" ]	en	trp	true
absorb	[ "bisingo kusubuisio" ]	en	trp	true
abstract	[ "abstract" ]	en	trp	true
absurd	[ "mwnwikothok" ]	en	trp	true
abundance	[ "khurumpui" ]	en	trp	true
abundant	[ "nangmani saibo bkwbang" ]	en	trp	true
abuse	[ "satokmung" ]	en	trp	true
accent	[ "kok samani goron" ]	en	trp	true
accept	[ "gosimani" ]	en	trp	true
access	[ "hablam" ]	en	trp	true
accident	[ "lahar patimung" ]	en	trp	true
accommodation	[ "tongjaknai thiti khlaimani" ]	en	trp	true
accomplish	[ "mungsaswk khlaimanmung" ]	en	trp	true
accomplished	[ "suargnang", "kwrwng" ]	en	trp	true
according	[ "hainohai" ]	en	trp	true
account	[ "lekhama" ]	en	trp	true
accuracy	[ "kwkwrwk" ]	en	trp	true
accurate	[ "kosol" ]	en	trp	true
accused	[ "majra berjak" ]	en	trp	true
ache	[ "samani" ]	en	trp	true
achieve	[ "manmung" ]	en	trp	true
achievement	[ "manmung" ]	en	trp	true
acknowledge	[ "gosimung rwmani" ]	en	trp	true
acquaintance	[ "sinilaimani" ]	en	trp	true
acquire	[ "manmani" ]	en	trp	true
acquired	[ "manjak" ]	en	trp	true
acquisition	[ "paijak manwi" ]	en	trp	true
across	[ "gwnangwi" ]	en	trp	true
act	[ "toke mwsamani" ]	en	trp	true
action	[ "yapri" ]	en	trp	true
active	[ "kwsrang" ]	en	trp	true
activities	[ "samung bwtangrok" ]	en	trp	true
activity	[ "samung khlaimanirok" ]	en	trp	true
actor	[ "toke mwsanaisa" ]	en	trp	true
actual	[ "kubuio" ]	en	trp	true
actually	[ "kubui sana thangkhe" ]	en	trp	true
acute	[ "belaikheno" ]	en	trp	true
adapt	[ "gwdalwi namani" ]	en	trp	true
add	[ "thepamani" ]	en	trp	true
addicted	[ "loi yakarmaya wngmani" ]	en	trp	true
addition	[ "kwtharwmani" ]	en	trp	true
additional	[ "teibo" ]	en	trp	true
address	[ "nikuma" ]	en	trp	true
adequate	[ "kulokmani" ]	en	trp	true
adjacent	[ "nanglaijak" ]	en	trp	true
adjective	[ "adjective" ]	en	trp	true
adjust	[ "jora bai gwdalwi tongma" ]	en	trp	true
admire	[ "naithokjakmani" ]	en	trp	true
admit	[ "damani" ]	en	trp	true
adopt	[ "dottok naharmani" ]	en	trp	true
adorable	[ "maya mannani chukjak" ]	en	trp	true
advance	[ "swkang bariya" ]	en	trp	true
advanced	[ "swkangjak" ]	en	trp	true
advantage	[ "sep kaham" ]	en	trp	true
adventure	[ "adventure" ]	en	trp	true
adverb	[ "Samungni dal" ]	en	trp	true
advertisement	[ "awaimung" ]	en	trp	true
advertising	[ "awai khlaimung" ]	en	trp	true
advice	[ "dagimung" ]	en	trp	true
advise	[ "kokbiti" ]	en	trp	true
advocate	[ "advocate" ]	en	trp	true
aesthetic	[ "naithothok" ]	en	trp	true
affair	[ "patimung" ]	en	trp	true
affect	[ "mari kwlaimani" ]	en	trp	true
affection	[ "maya" ]	en	trp	true
afford	[ "rang kwrwini bagwi mungsaswk paimaya wngmani" ]	en	trp	true
affordable	[ "pai manthokmani" ]	en	trp	true
afraid	[ "kirijak" ]	en	trp	true
after	[ "ulo" ]	en	trp	true
afternoon	[ "dipor" ]	en	trp	true
afterwards	[ "yagulo" ]	en	trp	true
again	[ "teisa" ]	en	trp	true
against	[ "kebengwi" ]	en	trp	true
age	[ "umor" ]	en	trp	true
agency	[ "khajak kaisa motha eba baniya" ]	en	trp	true
agent	[ "agent" ]	en	trp	true
aggressive	[ "jolikwrak" ]	en	trp	true
agile	[ "Kwsrang" ]	en	trp	true
ago	[ "swkango" ]	en	trp	true
agree	[ "gosimani" ]	en	trp	true
agreement	[ "gosimung bwlai" ]	en	trp	true
agriculture	[ "hukbar" ]	en	trp	true
ahead	[ "bwskango" ]	en	trp	true
aid	[ "chubachu" ]	en	trp	true
aim	[ "naharmung" ]	en	trp	true
air	[ "nobar" ]	en	trp	true

End of preview.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

Kokborok Digitalisation Project

The Kokborok Digitalisation Project is an initiative to curate and enhance parallel data for the Kokborok-English language pair. This project builds upon the SMOL dataset by Google, available on Hugging Face, and involves modifying and correcting it to better reflect the nuances of the local Kokborok dialect.

From the Author

"Language is a living, breathing entity—constantly evolving, shaping cultures, and connecting generations. When we digitally preserve and document it, we don’t just safeguard words; we ignite progress, making knowledge more accessible and enriching for all."

— Dr. Sudhamoy Debbarma

Overview

Objective: To collect, modify, and enhance data from various sources to create a high-quality, parallel dataset for the Kokborok-English language pair.
Dataset Origin: The dataset is derived from the SMOL dataset, which provides professionally translated data for 115 under-represented languages. This project adapts that dataset to serve the needs of the Kokborok-speaking community.

Dataset

The dataset consists of parallel translations in Kokborok-English.
It is a curated and improved version of the SMOL dataset.
Data corrections are based on local linguistic expertise and dialectal accuracy.

Usage

The dataset can be used for machine translation, linguistic research, and language preservation efforts.
Specific instructions on how to access or download the dataset will be added soon.

Contributing

Contributions are welcome! If you wish to contribute corrections or additional data, please open an issue or submit a pull request.

License

This project is licensed under the MIT License.

Acknowledgements

We thank the authors of the SMOL dataset and the related research work, whose contributions have significantly influenced this project. Special thanks to all contributors supporting under-represented languages and dialects.

Citation

If you use this work or any of its components in your research, please cite the following references:

@misc{caswell2025smol,
      title={SMOL: Professionally translated parallel data for 115 under-represented languages}, 
      author={Isaac Caswell and Elizabeth Nielsen and Jiaming Luo and Colin Cherry and Geza Kovacs and Hadar Shemtov and Partha Talukdar and Dinesh Tewari and Baba Mamadi Diane and Koulako Moussa Doumbouya and Djibrila Diane and Solo Farabado Cissé},
      year={2025},
      eprint={2502.12301},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2502.12301}, 
}

@misc{jones2023gatitos,
      title={Bilex Rx: Lexical Data Augmentation for Massively Multilingual Machine Translation}, 
      author={Alex Jones and Isaac Caswell and Ishank Saxena and Orhan Firat},
      year={2023},
      eprint={2303.15265},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2303.15265}, 
}

TODO

Gather phonetic data
Text to speech model

Downloads last month: 29