README.md

metadata

license: mit
dataset_info:
  features:
    - name: question
      dtype: string
    - name: table
      dtype: string
    - name: answer
      dtype: string
  splits:
    - name: train
      num_bytes: 7644946504
      num_examples: 847093
    - name: validation
      num_bytes: 8053382
      num_examples: 848
    - name: test
      num_bytes: 875174
      num_examples: 139
  download_size: 0
  dataset_size: 7653875060

Dataset Card for "hindiTabQA"

Usage

import pandas as pd
from datasets import load_dataset

hinditableQA = load_dataset("vaishali/hindiTabQA")

for sample in hinditableQA['train']:
  question = sample['question']
  input_table = pd.read_json(sample['table'], orient='split')
  answer = pd.read_json(sample['answer'], orient='split')

BibTeX entry and citation info

@misc{pal2024tablequestionansweringlowresourced,
      title={Table Question Answering for Low-resourced Indic Languages}, 
      author={Vaishali Pal and Evangelos Kanoulas and Andrew Yates and Maarten de Rijke},
      year={2024},
      eprint={2410.03576},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2410.03576}, 
}