how to evaluate models trained by bloom serires base model?

[Modas-Li]

please update and support this~

[jxhe]

maybe you can try evaluating through lm-evaluation-harness?

https://github.com/hkust-nlp/ceval#use-through-evaluation-harness