ai2_arc_with_ir
コレクションでコンテンツを整理
必要に応じて、コンテンツの保存と分類を行います。
高度な質問応答の研究を促進するために集められた、7,787 の本物の小学校レベルの多肢選択式科学問題の新しいデータセット。データセットは、チャレンジ セットとイージー セットに分割されます。前者には、検索ベースのアルゴリズムと単語共起アルゴリズムの両方によって間違って回答された質問のみが含まれます。また、タスクに関連する 1,400 万を超える科学文のコーパスと、このデータセットの 3 つのニューラル ベースライン モデルの実装も含まれています。私たちは ARC をコミュニティへの挑戦として提起します。
元のデータセットと比較して、これは UnifiedQA と同じ方法で情報検索によって得られたコンテキスト センテンスを追加します (参照: https://arxiv.org/abs/2005.00700 )。
FeaturesDict({
'answerKey': ClassLabel(shape=(), dtype=int64, num_classes=5),
'choices': Sequence({
'label': ClassLabel(shape=(), dtype=int64, num_classes=5),
'text': Text(shape=(), dtype=string),
}),
'id': Text(shape=(), dtype=string),
'paragraph': Text(shape=(), dtype=string),
'question': Text(shape=(), dtype=string),
})
特徴 | クラス | 形 | Dtype | 説明 |
---|
| 特徴辞書 | | | |
解答 | クラスラベル | | int64 | |
選択肢 | 順序 | | | |
選択肢/ラベル | クラスラベル | | int64 | |
選択肢/テキスト | 文章 | | ストリング | |
ID | 文章 | | ストリング | |
段落 | 文章 | | ストリング | |
質問 | 文章 | | ストリング | |
@article{allenai:arc,
author = {Peter Clark and Isaac Cowhey and Oren Etzioni and Tushar Khot and
Ashish Sabharwal and Carissa Schoenick and Oyvind Tafjord},
title = {Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge},
journal = {arXiv:1803.05457v1},
year = {2018},
}
@article{2020unifiedqa,
title={UnifiedQA: Crossing Format Boundaries With a Single QA System},
author={D. Khashabi and S. Min and T. Khot and A. Sabhwaral and O. Tafjord and P. Clark and H. Hajishirzi},
journal={arXiv preprint},
year={2020}
}
ai2_arc_with_ir/ARC-Challenge-IR (デフォルト構成)
スプリット | 例 |
---|
'test' | 1,172 |
'train' | 1,119 |
'validation' | 299 |
ai2_arc_with_ir/ARC-Easy-IR
スプリット | 例 |
---|
'test' | 2,376 |
'train' | 2,251 |
'validation' | 570 |
特に記載のない限り、このページのコンテンツはクリエイティブ・コモンズの表示 4.0 ライセンスにより使用許諾されます。コードサンプルは Apache 2.0 ライセンスにより使用許諾されます。詳しくは、Google Developers サイトのポリシーをご覧ください。Java は Oracle および関連会社の登録商標です。
最終更新日 2023-11-29 UTC。
[null,null,["最終更新日 2023-11-29 UTC。"],[],[],null,["# ai2_arc_with_ir\n\n\u003cbr /\u003e\n\n- **Description**:\n\nA new dataset of 7,787 genuine grade-school level, multiple-choice science\nquestions, assembled to encourage research in advanced question-answering. The\ndataset is partitioned into a Challenge Set and an Easy Set, where the former\ncontains only questions answered incorrectly by both a retrieval-based algorithm\nand a word co-occurrence algorithm. We are also including a corpus of over 14\nmillion science sentences relevant to the task, and an implementation of three\nneural baseline models for this dataset. We pose ARC as a challenge to the\ncommunity.\n\nCompared to the original dataset, this adds context sentences obtained through\ninformation retrieval in the same way as UnifiedQA (see:\n\u003chttps://arxiv.org/abs/2005.00700\u003e ).\n\n- **Additional Documentation** :\n [Explore on Papers With Code\n north_east](https://paperswithcode.com/dataset/arc)\n\n- **Homepage** : \u003chttps://allenai.org/data/arc\u003e\n\n- **Source code** :\n [`tfds.datasets.ai2_arc_with_ir.Builder`](https://github.com/tensorflow/datasets/tree/master/tensorflow_datasets/datasets/ai2_arc_with_ir/ai2_arc_with_ir_dataset_builder.py)\n\n- **Versions**:\n\n - **`1.0.0`** (default): No release notes.\n- **Download size** : `3.68 MiB`\n\n- **Auto-cached**\n ([documentation](https://www.tensorflow.org/datasets/performances#auto-caching)):\n Yes\n\n- **Feature structure**:\n\n FeaturesDict({\n 'answerKey': ClassLabel(shape=(), dtype=int64, num_classes=5),\n 'choices': Sequence({\n 'label': ClassLabel(shape=(), dtype=int64, num_classes=5),\n 'text': Text(shape=(), dtype=string),\n }),\n 'id': Text(shape=(), dtype=string),\n 'paragraph': Text(shape=(), dtype=string),\n 'question': Text(shape=(), dtype=string),\n })\n\n- **Feature documentation**:\n\n| Feature | Class | Shape | Dtype | Description |\n|---------------|--------------|-------|--------|-------------|\n| | FeaturesDict | | | |\n| answerKey | ClassLabel | | int64 | |\n| choices | Sequence | | | |\n| choices/label | ClassLabel | | int64 | |\n| choices/text | Text | | string | |\n| id | Text | | string | |\n| paragraph | Text | | string | |\n| question | Text | | string | |\n\n- **Supervised keys** (See\n [`as_supervised` doc](https://www.tensorflow.org/datasets/api_docs/python/tfds/load#args)):\n `None`\n\n- **Figure**\n ([tfds.show_examples](https://www.tensorflow.org/datasets/api_docs/python/tfds/visualization/show_examples)):\n Not supported.\n\n- **Citation**:\n\n @article{allenai:arc,\n author = {Peter Clark and Isaac Cowhey and Oren Etzioni and Tushar Khot and\n Ashish Sabharwal and Carissa Schoenick and Oyvind Tafjord},\n title = {Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge},\n journal = {arXiv:1803.05457v1},\n year = {2018},\n }\n @article{2020unifiedqa,\n title={UnifiedQA: Crossing Format Boundaries With a Single QA System},\n author={D. Khashabi and S. Min and T. Khot and A. Sabhwaral and O. Tafjord and P. Clark and H. Hajishirzi},\n journal={arXiv preprint},\n year={2020}\n }\n\nai2_arc_with_ir/ARC-Challenge-IR (default config)\n-------------------------------------------------\n\n- **Config description**: Challenge Set of 2590 \"hard\" questions (those that\n both a retrieval and a co-occurrence method fail to answer correctly)\n\n- **Dataset size** : `3.76 MiB`\n\n- **Splits**:\n\n| Split | Examples |\n|----------------|----------|\n| `'test'` | 1,172 |\n| `'train'` | 1,119 |\n| `'validation'` | 299 |\n\n- **Examples** ([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)):\n\nDisplay examples... \n\nai2_arc_with_ir/ARC-Easy-IR\n---------------------------\n\n- **Config description**: Easy Set of 5197 questions for the ARC Challenge.\n\n- **Dataset size** : `7.49 MiB`\n\n- **Splits**:\n\n| Split | Examples |\n|----------------|----------|\n| `'test'` | 2,376 |\n| `'train'` | 2,251 |\n| `'validation'` | 570 |\n\n- **Examples** ([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)):\n\nDisplay examples..."]]