search

mtbench101 / mt-bench-101

[ACL 2024] MT-Bench-101: A Fine-Grained Benchmark for Evaluating Large Language Models in Multi-Turn Dialogues
Apache License 2.0
48 stars 16 forks source link

论文都发表了，写在论文里，github仓库为空 #2

Closed victorjiax closed 5 months ago

victorjiax commented 7 months ago

能不能有相关的中文多轮评测数据，中文太少了

sefira commented 5 months ago

Thank you all for your interest in MT-Bench-101! We are currently organizing the code and data. In the meantime, you can contact xingyuanbu@gmail.com to obtain the beta version, which can be run on an OpenCompass-based code branch. Thank you!