README.md

metadata

dataset_info:
  - config_name: deduped_output
    features:
      - name: text
        dtype: string
      - name: id
        dtype: string
      - name: metadata
        dtype: string
    splits:
      - name: train
        num_bytes: 314811194363
        num_examples: 10030761
    download_size: 314811194363
    dataset_size: 1259244777452
configs:
  - config_name: deduped_output
    data_files:
      - split: train
        path: data/*.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="deduped_output", split="train", streaming=True)

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

License

Internal testing only

Statistics

	num_bytes	num_examples
deduped_output	314811194363	10030761