Afaan Oromo-Language-Model

Overview

This repository provides a machine translation model that translates between English and Afaan Oromo. The model is based on the Hugging Face transformers library and uses pre-trained models that are fine-tuned on English ↔ Afaan Oromo datasets.

Key Features:

• Translation between English and Afaan Oromo
• Fine-tuning of pre-trained models for improved accuracy
• Preprocessing and tokenization for datasets
• Easy-to-use interface for translation tasks

---

Installation

To get started with this project, follow the steps below:

1. Clone this repository:

   git clone https://github.com/your-username/afaan-oromo-translation.git
   cd afaan-oromo-translation
   
    Create and activate a Python virtual environment:
   

python -m venv venv source venv/bin/activate # For Linux/Mac venv\Scripts\activate # For Windows

Install the required libraries:

pip install -r requirements.txt

How to Use Translating Text:

You can use the pre-trained model for quick translation:

from transformers import MarianMTModel, MarianTokenizer

Load the pre-trained model and tokenizer

model_name = "Helsinki-NLP/opus-mt-en-orm" tokenizer = MarianTokenizer.from_pretrained(model_name) model = MarianMTModel.from_pretrained(model_name)

Translate text

text = "Hello, how are you?" inputs = tokenizer(text, return_tensors="pt") translated = model.generate(**inputs) output = tokenizer.decode(translated[0], skip_special_tokens=True)

print(output) # Output: Afaan Oromo translation

Fine-tuning the Model:

If you want to fine-tune the model with your custom dataset, follow these steps:

Prepare your English ↔ Afaan Oromo dataset in CSV format (with columns english_text and afaan_oromo_text).

Load and preprocess the dataset:

from datasets import load_dataset

dataset = load_dataset("csv", data_files={"train": "path/to/your_dataset.csv"}) train_data = dataset["train"]

Train the model:

from transformers import Trainer, TrainingArguments

training_args = TrainingArguments(
    output_dir="./afaan_oromo_translator",
    evaluation_strategy="epoch",
    learning_rate=5e-5,
    per_device_train_batch_size=8,
    per_device_eval_batch_size=8,
    num_train_epochs=3,
    weight_decay=0.01,
    logging_dir="./logs",
)

trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=train_data,
    tokenizer=tokenizer,
)

trainer.train()

Contributing

Contributions are welcome! If you'd like to contribute to the project, feel free to fork the repository and submit a pull request. License

This project is licensed under the MIT License. Contact

If you have any questions, feel free to contact me:

Name: Adugna Gizaw
Email: [email protected]

Acknowledgements

Hugging Face: for providing pre-trained models and the Transformers library
TensorFlow & PyTorch: for the machine learning frameworks