Duplicate this leaderboard to initialize your own!
Calculate GPU requirements for running LLMs
Convert PaddleOCR models to ONNX format
Text-To-Speech (TTS) Evaluation using objective metrics.
Convert Stable Diffusion checkpoint to Diffusers and open a PR
Retrain models for new data at edge devices
Visualize model performance on function calling tasks
Track, rank and evaluate open LLMs and chatbots
View and submit machine learning model evaluations
Create demo spaces for models on Hugging Face
Compare audio representation models using benchmark results
Persian Text Embedding Benchmark
Convert Hugging Face models to OpenVINO format
The Example Leaderboard Template is a starting point for creating custom leaderboards to track and compare the performance of various models. It allows users to easily duplicate and customize their own leaderboard to evaluate and benchmark different models based on specific criteria. This template is particularly useful for model benchmarking, providing a structured way to view and submit evaluations of large language models (LLMs).
What is the purpose of the Example Leaderboard Template?
The template is designed to provide a foundation for creating custom leaderboards to benchmark and evaluate the performance of various models, particularly LLMs.
How do I add a new model to the leaderboard?
You can add a new model by duplicating the template and then inputting the model's name, description, and performance data across the defined metrics.
Can I customize the metrics in the leaderboard?
Yes, the template is fully customizable. You can modify or add new metrics to suit your specific requirements and use case.