README.md

---
language:
- so
license: cc-by-4.0
task_categories:
- translation
pretty_name: Maay ↔ Maxaa Somali Parallel Translation Corpus
configs:
- config_name: default
  data_files:
  - split: train
    path: "data/train.jsonl"
  default: true
dataset_info:
  features:
  - name: source
    dtype: string
  - name: target
    dtype: string
  - name: source_lang
    dtype: string
  - name: target_lang
    dtype: string
  splits:
  - name: train
  - name: validation
  - name: test
---

# Maay ↔ Maxaa Somali Parallel Translation Corpus

Official parallel dataset for **Maay (Maay Maay)** and **Maxaa-tiri (Standard Somali)** dialects, collected and curated by the community via **MaayMaxaa DataHub**.

## 📊 Dataset Statistics

| Metric | Value |
|--------|-------|
| **Total Sentences** | 8 |
| **Maay → Maxaa** | 4 |
| **Maxaa → Maay** | 4 |
| **Avg Source Length** | 16 chars |
| **Avg Target Length** | 16 chars |
| **Version** | 0.1.0 |

## 🌐 Languages & Dialects

- **Maay (Maay Maay)**: Southern Somali language/dialect spoken across central and southern Somalia.
- **Maxaa-tiri (Standard Somali)**: Official written standard Somali dialect.

## 🚀 Quickstart (Python)

```python
from datasets import load_dataset

# Load dataset from Hugging Face Hub
dataset = load_dataset("SomaliDatasets/maay-maxaa-translation")

# Inspect train split
print(dataset["train"][0])
# {'source': '...', 'target': '...', 'source_lang': 'maay', 'target_lang': 'maxaa'}
```

## 📁 Schema & Fields

- `source`: Source text in original dialect.
- `target`: Translated target text.
- `source_lang`: Language code (`maay` or `maxaa`).
- `target_lang`: Language code (`maay` or `maxaa`).

## 📜 License

This dataset is released under the **Creative Commons Attribution 4.0 International (CC-BY-4.0)** license for research, model training, and open development of Somali NLP.