metadata
dataset_info:
features:
- name: answers
struct:
- name: answer_start
sequence: 'null'
- name: text
sequence: string
- name: inputs
dtype: string
- name: targets
dtype: string
splits:
- name: all_docs_eval
num_bytes: 7125701
num_examples: 10925
- name: validation
num_bytes: 752802
num_examples: 6489
- name: train_qa
num_bytes: 697367
num_examples: 6000
- name: train_ic_qa
num_bytes: 4540536
num_examples: 6000
- name: train
num_bytes: 38313328
num_examples: 63692
- name: eval_ic_qa
num_bytes: 4906186
num_examples: 6489
- name: eval_recite_qa
num_bytes: 4912675
num_examples: 6489
- name: all_docs
num_bytes: 7126313
num_examples: 10925
- name: eval_qa
num_bytes: 752802
num_examples: 6489
- name: train_recite_qa
num_bytes: 4546536
num_examples: 6000
- name: first_permute_docs
num_bytes: 37615961
num_examples: 57692
download_size: 37058553
dataset_size: 111290207