Mushkil

This repository hosts the Mushkil model, an AraT5V2 model fine-tuned for the task of Arabic text diacritization. The model treats diacritization as a machine translation problem, translating from undiacritized to diacritized Arabic text. It has been trained and validated on a dataset of over 550,000 Arabic sentences, achieving a high BLEU score of 0.93, reflecting its effectiveness in producing accurate diacritizations.

Model Description

The Mushkil model is based on the AraT5V2 architecture, a powerful transformer-based model pre-trained for various Arabic language tasks. For the purpose of diacritization, this model has been further trained to understand the context and apply appropriate diacritics to Arabic texts, which is crucial for correct pronunciation and understanding in reading Arabic scripts.

Getting Started

Installation

To use the model, you need to install the transformers library from Hugging Face. You can install it using pip:

pip install transformers

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text2text-generation", model="riotu-lab/mushkil")

pipe('ادخل نصك العربي')