What are expected to submit for the leaderboard integration?

open-compass / MathBench

[ACL 2024 Findings] MathBench: A Comprehensive Multi-Level Difficulty Mathematics Evaluation Dataset

https://open-compass.github.io/MathBench/

Apache License 2.0

73 stars 1 forks source link

What are expected to submit for the leaderboard integration? #7

Closed zhimin-z closed 2 months ago

zhimin-z commented 5 months ago

liushz commented 2 months ago

Thank you for your attention to MathBench! For new open-source models, a prediction directory for MathBench is wanted. For closed-source models and API models that are not easily accessible, we will soon launch a Community Leaderboard where these models can be ranked on MathBench. If you have any model names that excel in math, we would be happy to test them ourselves.