---
dataset_info:
  features:
  - name: gold
    dtype: large_string
  - name: subtask
    dtype: large_string
  - name: resp_0
    dtype: large_string
  - name: resp_1
    dtype: large_string
  - name: resp_2
    dtype: large_string
  - name: resp_3
    dtype: large_string
  - name: resp_0_correct
    dtype: bool
  - name: resp_1_correct
    dtype: bool
  - name: resp_2_correct
    dtype: bool
  - name: resp_3_correct
    dtype: bool
  - name: keep_0_src
    dtype: large_string
  - name: keep_0
    dtype: large_string
  - name: keep_0_correct
    dtype: bool
  - name: keep_1_src
    dtype: large_string
  - name: keep_1
    dtype: large_string
  - name: keep_1_correct
    dtype: bool
  - name: paragraphs
    dtype: large_string
  - name: question
    dtype: large_string
  splits:
  - name: train
    num_bytes: 1475535748
    num_examples: 31650
  download_size: 829716806
  dataset_size: 1475535748
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---