---
language:
- bbj
- en
license: apache-2.0
tags:
- translation
datasets:
- ghomala
dataset_info:
  features:
  - name: translation
    struct:
    - name: bbj
      dtype: string
    - name: en
      dtype: string
  splits:
  - name: train
    num_bytes: 1666787.0848231267
    num_examples: 6309
  - name: validation
    num_bytes: 208447.45758843666
    num_examples: 789
  - name: test
    num_bytes: 208447.45758843666
    num_examples: 789
  download_size: 1238492
  dataset_size: 2083682.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---

# Ghomala-en Translation Dataset

## Dataset Description
This is a parallel corpus for machine translation between Ghomala and en.
The dataset contains aligned sentences from the Ghomala Bible text corpus.

- **Languages**: Ghomala (bbj) → en
- **Dataset Type**: Parallel Corpus
- **Size**: 7887 parallel sentences
- **Source**: Ghomala Bible text corpus
- **License**: Apache 2.0

## Dataset Structure
- Format: Parallel text pairs
- Fields:
  - source_text: Ghomala text
  - target_text: en translation
- Splits:
  - Train: 80%
  - Validation: 10%
  - Test: 10%

## Usage
```python
from datasets import load_dataset

dataset = load_dataset("DS4H-ICTU/bbj-en-translation")
```

## Citation
``` 
@misc{ghomala_en_translation,
    title = {ghomala-en Translation Dataset},
    author = {NDE HURICH DILAN},
    year = 2024,
    publisher = Hugging Face
}
```

## Contact
For more information: ndedilan504@gmail.com