metadata
dataset_info:
features:
- name: id
dtype: string
- name: sequence
dtype: string
- name: name
dtype: string
- name: description
dtype: string
- name: features
dtype: int64
- name: seq_length
dtype: int64
- name: sequence_quality
dtype: float64
- name: text
dtype: string
- name: __index_level_0__
dtype: int64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 567591779
num_examples: 78918
download_size: 95799817
dataset_size: 567591779
configs:
- config_name: default
data_files:
- split: train
path: data/train-*