metadata
dataset_info:
features:
- name: text
dtype: string
- name: id
dtype: int64
- name: sentences
list:
- name: end_char
dtype: int64
- name: id
dtype: int64
- name: speaker
dtype: 'null'
- name: start_char
dtype: int64
- name: text
dtype: string
- name: tokens
list:
- name: deprel
dtype: string
- name: end_char
dtype: int64
- name: feats
dtype: string
- name: head
dtype: int64
- name: id
dtype: int64
- name: lemma
dtype: string
- name: start_char
dtype: int64
- name: text
dtype: string
- name: upos
dtype: string
- name: xpos
dtype: string
- name: coref_chains
sequence:
sequence:
sequence: int64
- name: genre
dtype: string
- name: meta_data
struct:
- name: comment
dtype: string
splits:
- name: train
num_bytes: 1558989
num_examples: 554
- name: validation
num_bytes: 384085
num_examples: 104
- name: test
num_bytes: 545213
num_examples: 146
download_size: 332859
dataset_size: 2488287
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
This dataset was generated by reformatting coref-data/superglue_wsc_raw into the indiscrim coreference format. See that repo for dataset details.
See ianporada/coref-data for additional conversion details and the conversion script.
Please create an issue in the repo above or in this dataset repo for any questions.