ABOUT_TEXT = ABOUT_TEXT = """
# 📝 About AfroBench Leaderboard

The **AfroBench Leaderboard** is a platform for evaluating multilingual language models across **64 African languages** and over **15 diverse NLP tasks**. These tasks span **classification**, **reasoning**, **question answering**, **summarization**, and **machine translation**, and are grounded in over **22 benchmark datasets** focused on low-resource and underrepresented languages.

The goal of this leaderboard is to:
- 🌍 Highlight the performance of LLMs on African languages.
- 🧪 Support diagnostic and task-level evaluation across different LLMs.
- ⚖️ Enable fair comparisons between open-source and closed models using both full and lite subsets of the benchmark.

This leaderboard supports two main views:
- **AfroBench**: The full evaluation benchmark organized by task, subtask, and dataset.
- **AfroBench-Lite**: A lightweight subset of the benchmark with a consistent set of languages across tasks, designed for efficient evaluation.

Each score is computed as the average across all selected columns and views, allowing flexible filtering and analysis.

---

## 🔗 More Information

To learn more about the benchmark, datasets, task definitions, and evaluation procedures, please visit the official project site:  
👉 [AfroBench Website](https://mcgill-nlp.github.io/AfroBench/index.html)

You can also explore:
- 📄 [AfroBench Paper on arXiv](https://arxiv.org/abs/2311.07978)  
- 🧑🏽‍💻 [AfroBench GitHub Repository](https://github.com/McGill-NLP/AfroBench)

"""


SUBMISSION_TEXT = """
Details on how to upload to teh huggingface leaderboard, COMING SOON!

In the meantime view how to run the evaluation in our [Github codebase](https://github.com/McGill-NLP/AfroBench) 
"""

SUBMISSION_TEXT_2 = """
"""
SUBMISSION_TEXT_3 = """
"""