THUDM / AgentBench

A Comprehensive Benchmark to Evaluate LLMs as Agents (ICLR'24)
https://llmbench.ai
Apache License 2.0
2.15k stars 150 forks source link

Leaderboard in machine readable format #2

Closed ogencoglu closed 1 year ago

ogencoglu commented 1 year ago

Thanks for the great work!

Can you provide the leaderboard results in some machine readable format (json, csv, xlsx etc.) in the repo as well?

xuyifan-0731 commented 1 year ago

Thanks for your advice.We have uploaded xlsx format in assets/agentbench_leaderboard.xlsx