metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: bool
- name: passage
dtype: string
- name: itv2 hi question
dtype: string
- name: itv2 hi passage
dtype: string
- name: itv2 hi question en
dtype: string
- name: itv2 hi passage en
dtype: string
splits:
- name: train
num_bytes: 25356277
num_examples: 9427
- name: validation
num_bytes: 8638084
num_examples: 3270
download_size: 16628479
dataset_size: 33994361
"boolq-translated"
Description: The BoolQ dataset is designed for natural language understanding tasks, particularly focused on boolean question answering. It contains questions paired with passages, where each question is associated with a boolean answer (true or false). The dataset is aimed at assessing a model's ability to comprehend and answer boolean questions based on given passages.
Dataset Info:
Features:
- Question: The boolean question presented to the model.
- Answer: The boolean answer (true or false) to the corresponding question.
- Passage: The passage containing information relevant to answering the question.
- Indic Translated Variants:
- Question (Hindi): Translated version of the question into Hindi.
- Passage (Hindi): Translated version of the passage into Hindi.
- Question (English): English version of the question.
- Passage (English): English version of the passage.
Splits:
- Train:
- Number of Examples: 9427
- Validation:
- Number of Examples: 3270
Download Size: Approximately 16.63 MB
Dataset Size: Approximately 33.99 MB