Awesome

Open LLM Leaderboard Scraper

The Open LLM Leaderboard Scraper is a Python script that allows you to retrieve and export data from the Open LLM Leaderboard.

[!WARNING] With the new update, the Open LLM Leaderboard has switched to a brand-new UI with a new frontend. Therefore, the scraping option is now unavailable. The old scraping method is still available in the old-scraping-method branch.

[!IMPORTANT] This repository still works by retrieving the dataframe from the official contents dataset.

Introduction
Prerequisites
Usage
Export Options
Output Files

Prerequisites

Before using the scraper, ensure you have the following prerequisites installed:

Python 3.x
Required Python packages

Usage

To use the Open LLM Leaderboard Scraper, follow these steps:

Clone the repository or download the script (main.py) to your local machine.

git clone https://github.com/Weyaxi/scrape-open-llm-leaderboard

Open a terminal or command prompt and navigate to the script's directory.

cd scrape-open-llm-leaderboard

Install the required packages using this command:

pip3 install -r requirements.txt

Run the script using the following command:

python3 main.py [options]

Export Options

The script supports three export options:

-csv: Export data to a CSV file.
-html: Export data to an HTML file.
-json: Export data to a JSON file.

You can use these options to specify the desired export format(s). For example, to export data in both CSV and HTML formats, run the following command:

python3 main.py -csv -html

If no export options are provided, the script will default to exporting data in CSV format.

Output Files

The scraper will generate one or more output files based on the export options selected. The output files will be named as follows:

open-llm-leaderboard.csv: Contains the scraped data in CSV format.
open-llm-leaderboard.html: Contains the scraped data in HTML format.
open-llm-leaderboard.json: Contains the scraped data in JSON format.

You can find these files in the same directory where you run the script.

Contributors

Special thanks to the following individuals who have contributed to this project: