File size: 1,211 Bytes
40349b3 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 |
import re
from datasets import load_dataset, Dataset, DatasetDict
def main():
for config in ['high', 'middle']:
datasets = load_dataset('ehovy/race', config)
_datasets = {}
for split in ['train', 'validation', 'test']:
data = []
for example in datasets[split]:
context = str(example['article'])
question = str(example['question'])
choices = example['options']
answer = example['answer']
answer_index = ord(answer) - ord('A')
data.append({
'context': context.strip(),
'question': question.strip(),
'choices': choices,
'answer': answer.strip(),
'answer_index': answer_index
})
dataset = Dataset.from_list(data)
_datasets[split] = dataset
datasets = DatasetDict({
'train': _datasets['train'],
'validation': _datasets['validation'],
'test': _datasets['test'],
})
datasets.push_to_hub('extraordinarylab/race', config)
if __name__ == '__main__':
main() |