metadata

license: bsd-2-clause
task_categories:
  - text-generation
language:
  - en
size_categories:
  - 1M<n<10M
configs:
  - config_name: default
    data_files:
      - split: train
        path: shakespeare.csv

Dataset Card for Dataset Name

This dataset is a part of the LEAF benchmark. The Shakespeare dataset is built from The Complete Works of William Shakespeare with the goal of the next character prediction.

Dataset Details

Dataset Description

Each sample is comprised of a text of 80 characters (x) and a next character (y).

Curated by: LEAF
Language(s) (NLP): English
License: BSD 2-Clause License

Dataset Sources

The code from the original repository was adopted to post it here.

Repository: https://github.com/TalwalkarLab/leaf
Paper: https://arxiv.org/abs/1812.01097

Uses

This dataset is intended to be used in Federated Learning settings. A pair of a character and a play denotes a unique user in the federation.

Direct Use

This dataset is designed to be used in FL settings. We recommend using Flower Dataset (flwr-datasets) and Flower (flwr).

To partition the dataset, do the following.

Install the package.

pip install flwr-datasets

Use the HF Dataset under the hood in Flower Datasets.

from flwr_datasets import FederatedDataset
from flwr_datasets.partitioner import NaturalIdPartitioner

fds = FederatedDataset(
    dataset="flwrlabs/shakespeare",
    partitioners={"train": NaturalIdPartitioner(partition_by="character_id")}
)
partition = fds.load_partition(node_id=0)

Dataset Structure

The dataset contains only train split. The split in the paper happens at each node only (no centralized dataset). The dataset is comprised of columns:

character_id: str - id denoting a pair of character + play (node in federated learning settings)
x: str - text of 80 characters
y: str - single character following the x

Please note that the data is temporal. Therefore, caution is needed when dividing it so as not to leak the information from the train set.

Dataset Creation

Curation Rationale

This dataset was created as a part of the LEAF benchmark.

Source Data

The Complete Works of William Shakespeare

Data Collection and Processing

For the preprocessing details, please refer to the original paper and the source code.

Who are the source data producers?

William Shakespeare

Citation

When working on the LEAF benchmark, please cite the original paper. If you're using this dataset with Flower Datasets, you can cite Flower.

BibTeX:

@article{DBLP:journals/corr/abs-1812-01097,
  author       = {Sebastian Caldas and
                  Peter Wu and
                  Tian Li and
                  Jakub Kone{\v{c}}n{\'y} and
                  H. Brendan McMahan and
                  Virginia Smith and
                  Ameet Talwalkar},
  title        = {{LEAF:} {A} Benchmark for Federated Settings},
  journal      = {CoRR},
  volume       = {abs/1812.01097},
  year         = {2018},
  url          = {http://arxiv.org/abs/1812.01097},
  eprinttype    = {arXiv},
  eprint       = {1812.01097},
  timestamp    = {Wed, 23 Dec 2020 09:35:18 +0100},
  biburl       = {https://dblp.org/rec/journals/corr/abs-1812-01097.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}

@article{DBLP:journals/corr/abs-2007-14390,
  author       = {Daniel J. Beutel and
                  Taner Topal and
                  Akhil Mathur and
                  Xinchi Qiu and
                  Titouan Parcollet and
                  Nicholas D. Lane},
  title        = {Flower: {A} Friendly Federated Learning Research Framework},
  journal      = {CoRR},
  volume       = {abs/2007.14390},
  year         = {2020},
  url          = {https://arxiv.org/abs/2007.14390},
  eprinttype    = {arXiv},
  eprint       = {2007.14390},
  timestamp    = {Mon, 03 Aug 2020 14:32:13 +0200},
  biburl       = {https://dblp.org/rec/journals/corr/abs-2007-14390.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}

Dataset Card Contact

In case of any doubts, please contact Flower Labs.