metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*
dataset_info:
features:
- name: source
dtype: string
- name: source_labels
dtype: string
- name: rouge_scores
dtype: string
- name: paper_id
dtype: string
- name: target
dtype: string
- name: full_source_text
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 17340567
num_examples: 1992
- name: test
num_bytes: 5620222
num_examples: 618
- name: validation
num_bytes: 5534448
num_examples: 619
download_size: 6328102
dataset_size: 28495237