metadata
dataset_info:
features:
- name: predicted_token
dtype: int64
- name: token_00
dtype: int64
- name: token_01
dtype: int64
- name: token_02
dtype: int64
- name: token_03
dtype: int64
- name: token_04
dtype: int64
- name: token_05
dtype: int64
- name: token_06
dtype: int64
- name: token_07
dtype: int64
- name: token_08
dtype: int64
- name: token_09
dtype: int64
- name: token_10
dtype: int64
- name: token_11
dtype: int64
- name: token_12
dtype: int64
- name: token_13
dtype: int64
- name: token_14
dtype: int64
- name: token_15
dtype: int64
- name: token_16
dtype: int64
- name: token_17
dtype: int64
- name: token_18
dtype: int64
- name: token_19
dtype: int64
splits:
- name: train
num_bytes: 13440000
num_examples: 80000
- name: test
num_bytes: 3360000
num_examples: 20000
download_size: 8171221
dataset_size: 16800000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*