Omar-youssef/english-egyptian-arabic-translator

English to Egyptian Arabic Neural Machine Translation Model

<!-- Provide a quick summary of what the model is/does. -->

This model translates English text into colloquial Egyptian Arabic. It is fine-tuned from the Helsinki-NLP/opus-mt-tc-big-en-ar base model on a specialized dataset of English-Egyptian Arabic translation pairs, ensuring natural, idiomatic, and culturally appropriate translations reflective of everyday Egyptian speech.

Model Details

Model Description

<!-- Provide a longer summary of what this model is. -->

This is a specialized neural machine translation model that translates from English to Egyptian Arabic (Masri/Egyptian Colloquial Arabic). Unlike standard Arabic translation models that produce Modern Standard Arabic (MSA), this model is trained to generate authentic colloquial Egyptian Arabic with natural speech patterns, slang, and tone appropriate to contemporary Egyptian communication.

• —Model type: Neural Machine Translation (Seq2Seq Transformer)
• —Language(s) (NLP): English (en), Egyptian Arabic (ar)
• —License: Apache 2.0
• —Finetuned from model: Helsinki-NLP/opus-mt-tc-big-en-ar

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

Direct Use

This model is designed for translating English content into colloquial Egyptian Arabic. Suitable applications include:

• —Social media content localization
• —Dialogue systems and chatbots
• —Educational materials adaptation
• —Content creation for Egyptian audiences
• —Travel and hospitality applications
• —Business communication in Egyptian contexts

Out-of-Scope Use

• —Formal or official document translation (prefer professional human translation or MSA models)
• —Legal or medical documents requiring precision and formal language
• —Highly technical or specialized terminology translation without domain-specific fine-tuning

How to Get Started with the Model

Use the code below to get started with the model:

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("Omar-youssef/english-egyptian-arabic-translator")
model = AutoModelForSeq2SeqLM.from_pretrained("Omar-youssef/english-egyptian-arabic-translator")

def translate(text):
    inputs = tokenizer.encode(text, return_tensors="pt")
    outputs = model.generate(inputs, num_beams=4, early_stopping=True)
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

print(translate("We haven't seen each other in a long time. We should catch up soon."))
# Output: إحنا مش بنشوف بعض من فترة طويلة. المفروض نلحق ببعضنا قريب.

print(translate("It's getting cold outside, so take a jacket with you."))
# Output: الجو طلع برد بره، فاخد جاكيت معاك.

Training Details

Training Data

The model is trained on the Omar-youssef/English-Egyptian-Arabic-Translation dataset, specifically designed for English-to-Egyptian Arabic translation. The training data includes:

Training Procedure

Training Hyperparameters

• —Training regime: (bf16)
• —Base Model: Helsinki-NLP/opus-mt-tc-big-en-ar
• —Learning Rate: [5e-6]
• —Batch Size: [16]
• —Epochs: [5]

Evaluation

<!-- This section describes the evaluation protocols and provides the results. -->

Testing Data, Factors & Metrics

Metrics

• —BLEU Score: [0.4111]

Technical Specifications

Model Architecture

• —Architecture: MarianMT (Encoder-Decoder Transformer)
• —Base: Helsinki-NLP/opus-mt-tc-big-en-ar
• —Parameters: ~500M
• —Task: Sequence-to-Sequence Translation

Compute Infrastructure

• —Hardware: Kaggle GPU (T4 )
• —Framework: Hugging Face Transformers
• —Training time: ~60 minutes

Citation

If you use this model, please cite the base model and dataset:

@misc{omar-youssef-egyptian-translator,
  author    = {Omar Youssef},
  title     = {English to Egyptian Arabic Translation Model},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/Omar-youssef/english-egyptian-arabic-translator}
}

Model Card Contact

For questions or feedback, open a discussion on this model's Hugging Face page.