wmt14_translate

  • 説明:

statmt.org のデータに基づいてデータセットを翻訳します。

複数のデータ ソースの組み合わせを使用して、さまざまな年のバージョンが存在します。ベースwmt_translateを使用すると、カスタムtfds.translate.wmt.WmtConfigを作成することにより、独自の構成を作成して独自のデータ/言語のペアを選択できます。

config = tfds.translate.wmt.WmtConfig(
    version="0.0.1",
    language_pair=("fr", "de"),
    subsets={
        tfds.Split.TRAIN: ["commoncrawl_frde"],
        tfds.Split.VALIDATION: ["euelections_dev2019"],
    },
)
builder = tfds.builder("wmt_translate", config=config)
  • 追加のドキュメント:コードを使用したペーパーの探索

  • ホームページ: http://www.statmt.org/wmt14/translation-task.html

  • ソース コード: tfds.translate.Wmt14Translate

  • バージョン:

    • 1.0.0 (デフォルト): リリース ノートはありません。
  • 手動ダウンロードの手順: このデータセットでは、ソース データを手動でdownload_config.manual_dir (デフォルトは~/tensorflow_datasets/downloads/manual/ ) にダウンロードする必要があります。
    ここにある wmt 構成の一部は、手動でダウンロードする必要があります。ダウンロードする必要がある正確なパス (およびファイル名) を確認するには、wmt.py を調べてください。

  • ( tfds.show_examples ): サポートされていません。

  • 引用

@InProceedings{bojar-EtAl:2014:W14-33,
  author    = {Bojar, Ondrej  and  Buck, Christian  and  Federmann, Christian  and  Haddow, Barry  and  Koehn, Philipp  and  Leveling, Johannes  and  Monz, Christof  and  Pecina, Pavel  and  Post, Matt  and  Saint-Amand, Herve  and  Soricut, Radu  and  Specia, Lucia  and  Tamchyna, Ale
{s} },
  title     = {Findings of the 2014 Workshop on Statistical Machine Translation},
  booktitle = {Proceedings of the Ninth Workshop on Statistical Machine Translation},
  month     = {June},
  year      = {2014},
  address   = {Baltimore, Maryland, USA},
  publisher = {Association for Computational Linguistics},
  pages     = {12--58},
  url       = {http://www.aclweb.org/anthology/W/W14/W14-3302}
}

wmt14_translate/cs-en (デフォルト設定)

  • 構成の説明: WMT 2014 cs-en 翻訳タスク データセット。

  • ダウンロードサイズ: 1.58 GiB

  • データセットサイズ: 2.89 GiB

  • 自動キャッシュ(ドキュメント): いいえ

  • スプリット:

スプリット
'test' 3,003
'train' 15,786,979
'validation' 3,000
  • 機能構造:
Translation({
    'cs': Text(shape=(), dtype=string),
    'en': Text(shape=(), dtype=string),
})
  • 機能のドキュメント:
特徴クラスDtype説明
翻訳
cs文章ストリング
ja文章ストリング

wmt14_translate/de-en

  • 構成の説明: WMT 2014 de-en 翻訳タスク データセット。

  • ダウンロードサイズ: 1.58 GiB

  • データセットサイズ: 1.36 GiB

  • 自動キャッシュ(ドキュメント): いいえ

  • スプリット:

スプリット
'test' 3,003
'train' 4,508,785
'validation' 3,000
  • 機能構造:
Translation({
    'de': Text(shape=(), dtype=string),
    'en': Text(shape=(), dtype=string),
})
  • 機能のドキュメント:
特徴クラスDtype説明
翻訳
文章ストリング
ja文章ストリング

wmt14_translate/fr-ja

  • 構成の説明: WMT 2014 fr-en 翻訳タスク データセット。

  • ダウンロードサイズ: 6.20 GiB

  • データセットサイズ: 14.65 GiB

  • 自動キャッシュ(ドキュメント): いいえ

  • スプリット:

スプリット
'test' 3,003
'train' 40,836,876
'validation' 3,000
  • 機能構造:
Translation({
    'en': Text(shape=(), dtype=string),
    'fr': Text(shape=(), dtype=string),
})
  • 機能のドキュメント:
特徴クラスDtype説明
翻訳
ja文章ストリング
フランス文章ストリング

wmt14_translate/hi-en

  • 構成の説明: WMT 2014 hi-en 翻訳タスク データセット。

  • ダウンロードサイズ: 44.65 MiB

  • データセットのサイズ: 77.53 MiB

  • 自動キャッシュ(ドキュメント): はい

  • スプリット:

スプリット
'test' 2,507
'train' 313,748
'validation' 520
  • 機能構造:
Translation({
    'en': Text(shape=(), dtype=string),
    'hi': Text(shape=(), dtype=string),
})
  • 機能のドキュメント:
特徴クラスDtype説明
翻訳
ja文章ストリング
こんにちは文章ストリング

wmt14_translate/ru-ja

  • 構成の説明: WMT 2014 ru-en 翻訳タスク データセット。

  • ダウンロードサイズ: 998.38 MiB

  • データセットサイズ: 839.40 MiB

  • 自動キャッシュ(ドキュメント): いいえ

  • スプリット:

スプリット
'test' 3,003
'train' 2,486,965
'validation' 3,000
  • 機能構造:
Translation({
    'en': Text(shape=(), dtype=string),
    'ru': Text(shape=(), dtype=string),
})
  • 機能のドキュメント:
特徴クラスDtype説明
翻訳
ja文章ストリング
文章ストリング