metadata
dataset_info:
features:
- name: error_type
dtype: string
- name: upos
dtype: string
- name: lemma
dtype: string
- name: feature
dtype: string
- name: bad_sentence
dtype: string
- name: sentence_start_index
dtype: int64
- name: error_word
dtype: float64
- name: agree_word
dtype: float64
- name: error_depth
dtype: float64
- name: agree_depth
dtype: float64
- name: sentence_length
dtype: float64
splits:
- name: train
num_bytes: 218831391
num_examples: 722395
download_size: 33194850
dataset_size: 218831391
configs:
- config_name: default
data_files:
- split: train
path: data/train-*