---

language: 🌐 Arabic | license: apache-2.0 | datasets: 📚 ARabic Reasoning QA | task_categories: 🔍 Question Answering | task_ids: 🧠 Reasoning

---

ARabic Reasoning QA

Dataset Description

The ARabic Reasoning QA dataset consists of Arabic reasoning questions designed to test various levels of reasoning skills. It includes questions with varying complexity levels: beginner, intermediate, and advanced. This dataset is suitable for training and evaluating models on reasoning tasks in Arabic.

Dataset Structure 🏗️

The dataset is organized into the following splits:

• train: Training data.
• eval: Evaluation data.
• test: Testing data.

Data Fields

• Question: The reasoning question in Arabic.
• Answer: The integer answer to the question.

Example

Here are some examples of the dataset:

Example 1:

• Question: لدي أربعة كتب وثلاثة دفاتر. كم عدد الأشياء التي لدي؟
• Answer: 7

Example 2:

• Question: إذا كان لديك ثلاث سيارات، وأعطيت واحدة منها، كم سيارة ستبقى لديك؟
• Answer: 2

Example 3:

• Question: كم عدد الأصابع في يد واحدة؟
• Answer: 5

Usage

To use this dataset, you can load it via the Hugging Face datasets library. Here's an example of how to load the dataset:

from datasets import load_dataset

dataset = load_dataset('MohammedNasser/ARabic_Reasoning_QA')

License

This dataset is licensed under the Apache 2.0 License.

Citation

If you use this dataset in your research, please cite it as follows:

@article{arabic_reasoning_benchmarking,
  title={Benchmarking Arabic Reasoning Models: Utilizing the ARabic Reasoning QA Dataset},
  author={M. N. Gaber},
  year={2024},
  publisher={Hugging Face Hub},
  url={https://huggingface.co/MohammedNasser/ARabic_Reasoning_QA}
}

This version reflects the Apache 2.0 license and includes all relevant details for users to properly cite and use the dataset.