README.md

metadata

task_categories:
  - text-generation
language:
  - de
dataset_info:
  - config_name: de
    features:
      - name: text
        dtype: large_string
      - name: url
        dtype: large_string
      - name: domain
        dtype: large_string
      - name: tlsh
        dtype: large_string
      - name: text_len
        dtype: uint32
      - name: lang_identification_prob
        dtype: float32
      - name: harmful_pp
        dtype: float32
    splits:
      - name: '2023_23'
        num_bytes: 14898393429
        num_examples: 7886185
    download_size: 14898393429
    dataset_size: 59593573716
configs:
  - config_name: de
    data_files:
      - split: '2023_23'
        path: data/2023-23/de/*.parquet

Some data

Just testing

Usage

from datasets import load_dataset

# init dataset in streaming mode
ds = load_dataset(repo_id, name="de", split="2023_23", streaming=True)

# print the first item of the dataset
print(next(iter(ds)))

License

Internal testing only