AttributeError: 'EvaluateArguments' object has no attribute 'eval_language'

[slatter666]

When I try to execute the testing script, I get this error, I guess perheps it's because you forget to update the code or anything? Another question is that should I modify dataset_fp.json? cause it only contains part of the whole dataset

[david101-hunter]

No, You only add like this to here

    eval_language: str = field(
        default = 'en',
        metadata={"help": "Which language of datasets to evaluate on. default is en, must be one of (en, zh)"}
    )

when run, depend on you to choose en or zh.