❓ Questions and Help
I used Reference Free Model Unbabel/wmt23-cometkiwi-da-xxl
My source langauge is English, target language is Chinese.
However there is some cases that the source(also target) text contains both Chinese and English.
I want the model to test fixed language pair English->Chinsese.
Is there any method ? I check both document and comet-score --help, there is no such function explaination.