ZaloAI-Jaist/VMLU

How to show results after evaluated

Opened this issue · 1 comments

lh0x00 commented

For example, FastChat was includes show benchmark results after evaluated done but VMLU is missing it, ref.

Can you add a convention script to do it and related about evaluate such as: draw a chart to compare, show evaluate, ..etc.

dear @lh0x00, the updating process is not fully automated and still involved a lot of manual work. If possible, please send an email to us at contact@vmlu.ai including model details and evaluation script for faster process.