Hugging Face's logo

Datasets:
LirioSandro
/
KmbPtMT

Tasks:
Translation
Text Generation
Languages:
Kimbundu
Portuguese
Tags:
kimbundu
low-resource
african-languages
nmt
License:
Dataset card Files
xet
 Community
Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
source: string
target: string
context: string
kmb: null
pt: null
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 619
to
{'kmb': Value('string'), 'pt': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2406, in _iter_arrow
                  pa_table = cast_table_to_features(pa_table, self.features)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2280, in cast_table_to_features
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              source: string
              target: string
              context: string
              kmb: null
              pt: null
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 619
              to
              {'kmb': Value('string'), 'pt': Value('string')}
              because column names don't match

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.

Dataset Card for Kimbundu-Portuguese Translation

Dataset Summary

This dataset contains 18,100 parallel sentence pairs between Kimbundu (a Bantu language from Angola) and Portuguese. It is a valuable resource aimed at bridging the data gap for low-resource African languages, specifically curated for Neural Machine Translation (NMT) tasks.

Language Details

  • Kimbundu (kmb): One of Angola's national languages, carrying significant cultural and historical importance.
  • Portuguese (pt): The official language of Angola and the target language for this dataset.

Dataset Structure

The dataset is organized in translation pairs. Each entry consists of:

  • kmb: Source text in Kimbundu.
  • pt: Corresponding translation in Portuguese.

Example

kmb pt
Muthu ni Dikixi. O Homem e o Monstro.

Creation and Curatorship

  • Data Sources: Books, religious websites, news outlets, and manual collection.
  • Preprocessing: The data underwent a cleaning process including deduplication, character normalization, and orthographic correction.

Applications

This dataset is ideal for:

  1. Training Machine Translation models (Transformer, MarianMT, etc.).
  2. Fine-tuning pre-trained multilingual language models (such as mBART or mT5).
  3. Comparative linguistic studies between Bantu and Romance languages.

Limitations and Ethical Considerations

As is common with low-resource datasets:

  • Size: With 18.1k pairs, models may struggle to capture complex grammatical nuances without data augmentation or transfer learning techniques.
  • Domain: The source data is predominantly from a specific domain (religious), generalization to everyday conversation may be limited.
  • Due to the scarcity of digital data in the Kimbundu language, this work used diverse sources for the purposes of cultural preservation and non-profit linguistic research. The dataset is made available under a scientific use license, respecting the integrity of the original sources through data fragmentation.
  • Purpose: This dataset was compiled exclusively for the purposes of linguistic preservation, scientific research and development of Natural Language Processing (NLP) technologies for the Kimbundu language.
  • Copyright: This dataset contains fragments of texts from different sources (literary, musical works and information platforms). Original copyright belongs to their respective holders.
  • Responsibility: The author of this dataset does not claim ownership of third-party content. By downloading or using this data, you assume full responsibility for any copyright infringements that may arise from commercial use or improper redistribution.
  • Removal: If you own the rights to any content here and wish to have it removed, please contact lirio.ramalheira1@gmail.com.

How to Cite

If you use this dataset in your research, please use the following citation: Ramalheira, L., Costa, A., Rodrigues, C. V., & Chau, F. I. (2026). Improving Kimbundu–Portuguese Neural Machine Translation through Fine-Tuning of Multilingual Models (Version V1.0). Zenodo. https://doi.org/10.5281/zenodo.18890117
Downloads last month
32
Total file size:
1.99 MB

Models trained or fine-tuned on LirioSandro/KmbPtMT