Ibani is an indigenous language spoken primarily in Rivers State, Nigeria. Due to the lack of an official ISO-639-3 code, this project uses the custom language identifier ibani, consistently applied across the model and dataset.

Usage

python

from transformers import MarianMTModel, MarianTokenizer

model = MarianMTModel.from_pretrained("your-username/ibani-translator")
tokenizer = MarianTokenizer.from_pretrained("your-username/ibani-translator")

text = "I eat fish"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs)
translation = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(translation)

Training Data

Trained on English-Ibani parallel sentences from the ibani_eng dataset.

Performance

[Add your evaluation metrics here]

Model Details

Base Model: Helsinki-NLP/opus-mt-en-mul
Language Pair: English → Ibani
Task: Machine Translation
Framework: Hugging Face Transformers

Limitations

This is a low-resource language model
Performance may vary with complex sentence structures
Best results with simple to moderate complexity sentences

Citation

If you use this model, please cite:

bibtex

@misc{ibani-translator,
  author = {Your Name},
  title = {English to Ibani Translation Model},
  year = {2025},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/your-username/ibani-translator}}
}

williampepple1/ibani-translator

English to Ibani Translation Model

Language