I'm trying to understand if the model performs equally well when prompting in English and in Chinese, is there data on this?
I'm trying to understand if the model performs equally well when prompting in English and in Chinese, is there data on this?