metadata
dataset_info:
- config_name: eng_Latn
features:
- name: text
dtype: string
- name: language_score
dtype: string
- name: source_file_path
dtype: string
- name: source_zip_content_path
dtype: string
- name: doi
dtype: string
- name: authors
dtype: string
- name: title
dtype: string
- name: pdf_md5_hash
dtype: string
- name: publisher
dtype: string
- name: journal
dtype: string
- name: journal_abbrev
dtype: string
- name: publication_year
dtype: uint32
splits:
- name: train
num_bytes: 11190029452
num_examples: 919043
download_size: 11190029452
dataset_size: 44760117808
configs:
- config_name: eng_Latn
data_files:
- split: train
path: '*.parquet'
Some data name
Just testing... this dataset is temporary and will be removed soon.
Usage
from datasets import load_dataset
# init dataset in streaming mode
ds = load_dataset(repo_id, name="eng_Latn", split="train", streaming=True)
# print the first item of the dataset
print(next(iter(ds)))
License
Internal testing only
Statistics
num_bytes | num_examples | |
---|---|---|
eng_Latn | 11190029452 | 919043 |