data.py

import datasets
import tasksource
import inspect
import copy


def get_kwarg_names(func):
    return [k for k, v in inspect.signature(func).parameters.items() if v.default != v.empty]

class Dataset(datasets.GeneratorBasedBuilder):

    builder_configs = {}
    def __init__(self,*args,**kwargs):
        self.BUILDER_CONFIG_CLASS.__post_init__=lambda x:x

        base_kwargs_names=get_kwarg_names(super().__init__)
        gen_kwargs={}
        for k,v in copy.deepcopy(kwargs).items():
            if k not in base_kwargs_names:
                gen_kwargs[k]=v
                del kwargs[k]
        self.gen_kwargs=gen_kwargs
        return super().__init__(*args,**kwargs)
        
    def _info(self):
        return datasets.DatasetInfo()

    def _split_generators(self, _):
        return [datasets.SplitGenerator(name=name, gen_kwargs={"split_name": name}) for name in ['train','validation','test']]

    def _generate_examples(self, split_name):
        i=0
        for x in tasksource.load_task(self.config.name,**self.gen_kwargs)[split_name]:
            i+=1
            yield i,x

    def _download_and_prepare(self, dl_manager, verification_mode, **prepare_splits_kwargs):
        result = super()._download_and_prepare(dl_manager, "no_checks", **prepare_splits_kwargs)
        return result