unified_qa

  • Deskripsi :

Tolok ukur UnifiedQA terdiri dari 20 kumpulan data penjawab pertanyaan (QA) utama (masing-masing mungkin memiliki beberapa versi) yang menargetkan format yang berbeda serta berbagai fenomena linguistik yang kompleks. Kumpulan data ini dikelompokkan ke dalam beberapa format/kategori, antara lain: QA ekstraktif, QA abstraktif, QA pilihan ganda, dan QA ya/tidak. Selain itu, set kontras digunakan untuk beberapa set data (dilambangkan dengan " set kontras"). Set evaluasi ini adalah gangguan yang dibuat oleh pakar yang menyimpang dari pola yang umum dalam kumpulan data asli. Untuk beberapa kumpulan data yang tidak dilengkapi dengan paragraf bukti, dua varian disertakan: satu di mana kumpulan data digunakan apa adanya dan yang lain menggunakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan, yang ditandai dengan tag "_ir".

Informasi lebih lanjut dapat ditemukan di: https://github.com/alllenai/unifiedqa

FeaturesDict({
    'input': string,
    'output': string,
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
memasukkan Tensor rangkaian
keluaran Tensor rangkaian

unified_qa/ai2_science_elementary (konfigurasi default)

  • Deskripsi konfigurasi : Dataset AI2 Science Questions terdiri dari pertanyaan yang digunakan dalam penilaian siswa di Amerika Serikat untuk semua tingkat sekolah dasar dan menengah. Setiap pertanyaan adalah format pilihan ganda 4 arah dan mungkin menyertakan atau tidak menyertakan elemen diagram. Set ini terdiri dari soal-soal yang digunakan untuk tingkat kelas sekolah dasar.

  • Ukuran unduhan : 345.59 KiB

  • Ukuran dataset : 390.02 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 542
'train' 623
'validation' 123
  • Kutipan :
http://data.allenai.org/ai2-science-questions

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/ai2_science_middle

  • Deskripsi konfigurasi : Dataset AI2 Science Questions terdiri dari pertanyaan yang digunakan dalam penilaian siswa di Amerika Serikat untuk semua tingkat sekolah dasar dan menengah. Setiap pertanyaan adalah format pilihan ganda 4 arah dan mungkin menyertakan atau tidak menyertakan elemen diagram. Set ini terdiri dari pertanyaan yang digunakan untuk tingkat kelas sekolah menengah.

  • Ukuran unduhan : 428.41 KiB

  • Ukuran dataset : 477.40 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 679
'train' 605
'validation' 125
  • Kutipan :
http://data.allenai.org/ai2-science-questions

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/ambigqa

  • Deskripsi konfigurasi : AmbigQA adalah tugas menjawab pertanyaan domain terbuka yang melibatkan menemukan setiap jawaban yang masuk akal, dan kemudian menulis ulang pertanyaan untuk masing-masing jawaban untuk menyelesaikan ambiguitas.

  • Ukuran unduhan : 2.27 MiB

  • Ukuran dataset : 3.04 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 19.806
'validation' 5.674
  • Kutipan :
@inproceedings{min-etal-2020-ambigqa,
    title = "{A}mbig{QA}: Answering Ambiguous Open-domain Questions",
    author = "Min, Sewon  and
      Michael, Julian  and
      Hajishirzi, Hannaneh  and
      Zettlemoyer, Luke",
    booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.emnlp-main.466",
    doi = "10.18653/v1/2020.emnlp-main.466",
    pages = "5783--5797",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_easy

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "mudah".

  • Ukuran unduhan : 1.24 MiB

  • Ukuran dataset : 1.42 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 2.376
'train' 2.251
'validation' 570
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_easy_dev

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "mudah".

  • Ukuran unduhan : 1.24 MiB

  • Ukuran dataset : 1.42 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 2.376
'train' 2.251
'validation' 570
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_easy_with_ir

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "mudah". Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 7.00 MiB

  • Ukuran dataset : 7.17 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 2.376
'train' 2.251
'validation' 570
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_easy_with_ir_dev

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "mudah". Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 7.00 MiB

  • Ukuran dataset : 7.17 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 2.376
'train' 2.251
'validation' 570
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_hard

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "sulit".

  • Ukuran unduhan : 758.03 KiB

  • Ukuran dataset : 848.28 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.172
'train' 1.119
'validation' 299
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_hard_dev

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "sulit".

  • Ukuran unduhan : 758.03 KiB

  • Ukuran dataset : 848.28 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.172
'train' 1.119
'validation' 299
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_hard_with_ir

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "sulit". Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 3.53 MiB

  • Ukuran dataset : 3.62 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.172
'train' 1.119
'validation' 299
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/arc_hard_with_ir_dev

  • Deskripsi konfigurasi : Kumpulan data ini terdiri dari pertanyaan sains pilihan ganda tingkat sekolah dasar asli, yang dikumpulkan untuk mendorong penelitian dalam menjawab pertanyaan tingkat lanjut. Dataset dipartisi menjadi Kumpulan Tantangan dan Kumpulan Mudah, di mana yang pertama hanya berisi pertanyaan yang dijawab salah oleh algoritme berbasis pengambilan dan algoritme kejadian bersama kata. Set ini terdiri dari pertanyaan "sulit". Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 3.53 MiB

  • Ukuran dataset : 3.62 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.172
'train' 1.119
'validation' 299
  • Kutipan :
@article{clark2018think,
    title={Think you have solved question answering? try arc, the ai2 reasoning challenge},
    author={Clark, Peter and Cowhey, Isaac and Etzioni, Oren and Khot, Tushar and Sabharwal, Ashish and Schoenick, Carissa and Tafjord, Oyvind},
    journal={arXiv preprint arXiv:1803.05457},
    year={2018}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/boolq

  • Deskripsi konfigurasi : BoolQ adalah kumpulan data penjawab pertanyaan untuk pertanyaan ya/tidak. Pertanyaan-pertanyaan ini muncul secara alami ---dihasilkan dalam pengaturan yang tidak diminta dan tidak dibatasi. Setiap contoh adalah triplet dari (pertanyaan, bagian, jawaban), dengan judul halaman sebagai konteks tambahan opsional. Penyiapan klasifikasi pasangan teks mirip dengan tugas inferensi bahasa alami yang ada.

  • Ukuran unduhan : 7.77 MiB

  • Ukuran dataset : 8.20 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 9.427
'validation' 3.270
  • Kutipan :
@inproceedings{clark-etal-2019-boolq,
    title = "{B}ool{Q}: Exploring the Surprising Difficulty of Natural Yes/No Questions",
    author = "Clark, Christopher  and
      Lee, Kenton  and
      Chang, Ming-Wei  and
      Kwiatkowski, Tom  and
      Collins, Michael  and
      Toutanova, Kristina",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1300",
    doi = "10.18653/v1/N19-1300",
    pages = "2924--2936",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/boolq_np

  • Deskripsi konfigurasi : BoolQ adalah kumpulan data penjawab pertanyaan untuk pertanyaan ya/tidak. Pertanyaan-pertanyaan ini muncul secara alami ---dihasilkan dalam pengaturan yang tidak diminta dan tidak dibatasi. Setiap contoh adalah triplet dari (pertanyaan, bagian, jawaban), dengan judul halaman sebagai konteks tambahan opsional. Penyiapan klasifikasi pasangan teks mirip dengan tugas inferensi bahasa alami yang ada. Versi ini menambahkan gangguan alami ke versi aslinya.

  • Ukuran unduhan : 10.80 MiB

  • Ukuran dataset : 11.40 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 9.727
'validation' 7.596
  • Kutipan :
@inproceedings{khashabi-etal-2020-bang,
    title = "More Bang for Your Buck: Natural Perturbation for Robust Question Answering",
    author = "Khashabi, Daniel  and
      Khot, Tushar  and
      Sabharwal, Ashish",
    booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.emnlp-main.12",
    doi = "10.18653/v1/2020.emnlp-main.12",
    pages = "163--170",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/commonsenseqa

  • Deskripsi konfigurasi : CommonsenseQA adalah kumpulan data jawaban pertanyaan pilihan ganda baru yang memerlukan berbagai jenis pengetahuan akal sehat untuk memprediksi jawaban yang benar . Ini berisi pertanyaan dengan satu jawaban yang benar dan empat jawaban distraktor.

  • Ukuran unduhan : 1.79 MiB

  • Ukuran dataset : 2.19 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.140
'train' 9.741
'validation' 1.221
  • Kutipan :
@inproceedings{talmor-etal-2019-commonsenseqa,
    title = "{C}ommonsense{QA}: A Question Answering Challenge Targeting Commonsense Knowledge",
    author = "Talmor, Alon  and
      Herzig, Jonathan  and
      Lourie, Nicholas  and
      Berant, Jonathan",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1421",
    doi = "10.18653/v1/N19-1421",
    pages = "4149--4158",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/commonsenseqa_test

  • Deskripsi konfigurasi : CommonsenseQA adalah kumpulan data jawaban pertanyaan pilihan ganda baru yang memerlukan berbagai jenis pengetahuan akal sehat untuk memprediksi jawaban yang benar . Ini berisi pertanyaan dengan satu jawaban yang benar dan empat jawaban distraktor.

  • Ukuran unduhan : 1.79 MiB

  • Ukuran dataset : 2.19 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.140
'train' 9.741
'validation' 1.221
  • Kutipan :
@inproceedings{talmor-etal-2019-commonsenseqa,
    title = "{C}ommonsense{QA}: A Question Answering Challenge Targeting Commonsense Knowledge",
    author = "Talmor, Alon  and
      Herzig, Jonathan  and
      Lourie, Nicholas  and
      Berant, Jonathan",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1421",
    doi = "10.18653/v1/N19-1421",
    pages = "4149--4158",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/contrast_sets_boolq

  • Deskripsi konfigurasi : BoolQ adalah kumpulan data penjawab pertanyaan untuk pertanyaan ya/tidak. Pertanyaan-pertanyaan ini muncul secara alami ---dihasilkan dalam pengaturan yang tidak diminta dan tidak dibatasi. Setiap contoh adalah triplet dari (pertanyaan, bagian, jawaban), dengan judul halaman sebagai konteks tambahan opsional. Penyiapan klasifikasi pasangan teks mirip dengan tugas inferensi bahasa alami yang ada. Versi ini menggunakan set kontras. Set evaluasi ini adalah gangguan yang dibuat oleh pakar yang menyimpang dari pola yang umum dalam kumpulan data asli.

  • Ukuran unduhan : 438.51 KiB

  • Ukuran dataset : 462.35 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 340
'validation' 340
  • Kutipan :
@inproceedings{clark-etal-2019-boolq,
    title = "{B}ool{Q}: Exploring the Surprising Difficulty of Natural Yes/No Questions",
    author = "Clark, Christopher  and
      Lee, Kenton  and
      Chang, Ming-Wei  and
      Kwiatkowski, Tom  and
      Collins, Michael  and
      Toutanova, Kristina",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1300",
    doi = "10.18653/v1/N19-1300",
    pages = "2924--2936",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/contrast_sets_drop

  • Deskripsi konfigurasi : DROP adalah tolok ukur QA yang dibuat secara crowdsourced, di mana sistem harus menyelesaikan referensi dalam sebuah pertanyaan, mungkin ke beberapa posisi input, dan melakukan operasi diskrit terhadapnya (seperti penambahan, penghitungan, atau penyortiran). Operasi ini membutuhkan pemahaman yang jauh lebih komprehensif tentang isi paragraf daripada yang diperlukan untuk kumpulan data sebelumnya. Versi ini menggunakan set kontras. Set evaluasi ini adalah gangguan yang dibuat oleh pakar yang menyimpang dari pola yang umum dalam kumpulan data asli.

  • Ukuran unduhan : 2.20 MiB

  • Ukuran dataset : 2.26 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 947
'validation' 947
  • Kutipan :
@inproceedings{dua-etal-2019-drop,
    title = "{DROP}: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs",
    author = "Dua, Dheeru  and
      Wang, Yizhong  and
      Dasigi, Pradeep  and
      Stanovsky, Gabriel  and
      Singh, Sameer  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1246",
    doi = "10.18653/v1/N19-1246",
    pages = "2368--2378",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/contrast_sets_quoref

  • Deskripsi konfigurasi : Kumpulan data ini menguji kemampuan penalaran inti dari sistem pemahaman bacaan. Dalam tolok ukur pemilihan rentang ini yang berisi pertanyaan tentang paragraf dari Wikipedia, sistem harus menyelesaikan referensi keras sebelum memilih rentang yang sesuai dalam paragraf untuk menjawab pertanyaan. Versi ini menggunakan set kontras. Set evaluasi ini adalah gangguan yang dibuat oleh pakar yang menyimpang dari pola yang umum dalam kumpulan data asli.

  • Ukuran unduhan : 2.60 MiB

  • Ukuran dataset : 2.65 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 700
'validation' 700
  • Kutipan :
@inproceedings{dasigi-etal-2019-quoref,
    title = "{Q}uoref: A Reading Comprehension Dataset with Questions Requiring Coreferential Reasoning",
    author = "Dasigi, Pradeep  and
      Liu, Nelson F.  and
      Marasovi{'c}, Ana  and
      Smith, Noah A.  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-1606",
    doi = "10.18653/v1/D19-1606",
    pages = "5925--5932",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/contrast_sets_ropes

  • Deskripsi konfigurasi : Kumpulan data ini menguji kemampuan sistem untuk menerapkan pengetahuan dari bagian teks ke situasi baru. Suatu sistem disajikan dengan bagian latar belakang yang berisi hubungan kausal atau kualitatif (misalnya, "penyerbuk hewan meningkatkan efisiensi pembuahan pada bunga"), situasi baru yang menggunakan latar belakang ini, dan pertanyaan yang memerlukan penalaran tentang efek hubungan dalam bagian latar belakang dalam konteks situasi. Versi ini menggunakan set kontras. Set evaluasi ini adalah gangguan yang dibuat oleh pakar yang menyimpang dari pola yang umum dalam kumpulan data asli.

  • Ukuran unduhan : 1.97 MiB

  • Ukuran dataset : 2.04 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 974
'validation' 974
  • Kutipan :
@inproceedings{lin-etal-2019-reasoning,
    title = "Reasoning Over Paragraph Effects in Situations",
    author = "Lin, Kevin  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5808",
    doi = "10.18653/v1/D19-5808",
    pages = "58--62",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/drop

  • Deskripsi konfigurasi : DROP adalah tolok ukur QA yang dibuat secara crowdsourced, di mana sistem harus menyelesaikan referensi dalam sebuah pertanyaan, mungkin ke beberapa posisi input, dan melakukan operasi diskrit terhadapnya (seperti penambahan, penghitungan, atau penyortiran). Operasi ini membutuhkan pemahaman yang jauh lebih komprehensif tentang isi paragraf daripada yang diperlukan untuk kumpulan data sebelumnya.

  • Ukuran unduhan : 105.18 MiB

  • Ukuran dataset : 108.16 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 77.399
'validation' 9.536
  • Kutipan :
@inproceedings{dua-etal-2019-drop,
    title = "{DROP}: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs",
    author = "Dua, Dheeru  and
      Wang, Yizhong  and
      Dasigi, Pradeep  and
      Stanovsky, Gabriel  and
      Singh, Sameer  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N19-1246",
    doi = "10.18653/v1/N19-1246",
    pages = "2368--2378",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/mctest

  • Deskripsi konfigurasi : MCTest memerlukan mesin untuk menjawab pertanyaan pemahaman bacaan pilihan ganda tentang cerita fiksi, yang secara langsung menangani tujuan tingkat tinggi dari pemahaman mesin domain terbuka. Pemahaman membaca dapat menguji kemampuan tingkat lanjut seperti penalaran kausal dan memahami dunia, namun dengan menjadi pilihan ganda, tetap memberikan metrik yang jelas. Dengan menjadi fiksi, jawabannya biasanya hanya dapat ditemukan dalam cerita itu sendiri. Cerita dan pertanyaan juga dibatasi dengan hati-hati untuk yang dapat dipahami oleh anak kecil, sehingga mengurangi pengetahuan dunia yang diperlukan untuk tugas tersebut.

  • Ukuran unduhan : 2.14 MiB

  • Ukuran dataset : 2.20 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 1.480
'validation' 320
  • Kutipan :
@inproceedings{richardson-etal-2013-mctest,
    title = "{MCT}est: A Challenge Dataset for the Open-Domain Machine Comprehension of Text",
    author = "Richardson, Matthew  and
      Burges, Christopher J.C.  and
      Renshaw, Erin",
    booktitle = "Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing",
    month = oct,
    year = "2013",
    address = "Seattle, Washington, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D13-1020",
    pages = "193--203",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/mctest_corrected_the_separator

  • Deskripsi konfigurasi : MCTest memerlukan mesin untuk menjawab pertanyaan pemahaman bacaan pilihan ganda tentang cerita fiksi, yang secara langsung menangani tujuan tingkat tinggi dari pemahaman mesin domain terbuka. Pemahaman membaca dapat menguji kemampuan tingkat lanjut seperti penalaran kausal dan memahami dunia, namun dengan menjadi pilihan ganda, tetap memberikan metrik yang jelas. Dengan menjadi fiksi, jawabannya biasanya hanya dapat ditemukan dalam cerita itu sendiri. Cerita dan pertanyaan juga dibatasi dengan hati-hati untuk yang dapat dipahami oleh anak kecil, sehingga mengurangi pengetahuan dunia yang diperlukan untuk tugas tersebut.

  • Ukuran unduhan : 2.15 MiB

  • Ukuran dataset : 2.21 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 1.480
'validation' 320
  • Kutipan :
@inproceedings{richardson-etal-2013-mctest,
    title = "{MCT}est: A Challenge Dataset for the Open-Domain Machine Comprehension of Text",
    author = "Richardson, Matthew  and
      Burges, Christopher J.C.  and
      Renshaw, Erin",
    booktitle = "Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing",
    month = oct,
    year = "2013",
    address = "Seattle, Washington, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D13-1020",
    pages = "193--203",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/multirc

  • Deskripsi konfigurasi : MultiRC adalah tantangan pemahaman bacaan di mana pertanyaan hanya dapat dijawab dengan mempertimbangkan informasi akun dari beberapa kalimat. Pertanyaan dan jawaban untuk tantangan ini diajukan dan diverifikasi melalui eksperimen crowdsourcing 4 langkah. Kumpulan data berisi pertanyaan untuk paragraf di 7 domain yang berbeda (ilmu sekolah dasar, berita, panduan perjalanan, cerita fiksi, dll) membawa keragaman linguistik ke dalam teks dan kata-kata pertanyaan.

  • Ukuran unduhan : 897.09 KiB

  • Ukuran dataset : 918.42 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 312
'validation' 312
  • Kutipan :
@inproceedings{khashabi-etal-2018-looking,
    title = "Looking Beyond the Surface: A Challenge Set for Reading Comprehension over Multiple Sentences",
    author = "Khashabi, Daniel  and
      Chaturvedi, Snigdha  and
      Roth, Michael  and
      Upadhyay, Shyam  and
      Roth, Dan",
    booktitle = "Proceedings of the 2018 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers)",
    month = jun,
    year = "2018",
    address = "New Orleans, Louisiana",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/N18-1023",
    doi = "10.18653/v1/N18-1023",
    pages = "252--262",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/narrativeqa

  • Deskripsi konfigurasi : NarrativeQA adalah kumpulan data cerita berbahasa Inggris dan pertanyaan terkait yang dirancang untuk menguji pemahaman bacaan, terutama pada dokumen panjang.

  • Ukuran unduhan : 308.28 MiB

  • Ukuran dataset : 311.22 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'test' 21.114
'train' 65.494
'validation' 6.922
  • Kutipan :
@article{kocisky-etal-2018-narrativeqa,
    title = "The {N}arrative{QA} Reading Comprehension Challenge",
    author = "Ko{
{c} }isk{'y}, Tom{'a}{
{s} }  and
      Schwarz, Jonathan  and
      Blunsom, Phil  and
      Dyer, Chris  and
      Hermann, Karl Moritz  and
      Melis, G{'a}bor  and
      Grefenstette, Edward",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "6",
    year = "2018",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q18-1023",
    doi = "10.1162/tacl_a_00023",
    pages = "317--328",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/narrativeqa_dev

  • Deskripsi konfigurasi : NarrativeQA adalah kumpulan data cerita berbahasa Inggris dan pertanyaan terkait yang dirancang untuk menguji pemahaman bacaan, terutama pada dokumen panjang.

  • Ukuran unduhan : 308.28 MiB

  • Ukuran dataset : 311.22 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'test' 21.114
'train' 65.494
'validation' 6.922
  • Kutipan :
@article{kocisky-etal-2018-narrativeqa,
    title = "The {N}arrative{QA} Reading Comprehension Challenge",
    author = "Ko{
{c} }isk{'y}, Tom{'a}{
{s} }  and
      Schwarz, Jonathan  and
      Blunsom, Phil  and
      Dyer, Chris  and
      Hermann, Karl Moritz  and
      Melis, G{'a}bor  and
      Grefenstette, Edward",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "6",
    year = "2018",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q18-1023",
    doi = "10.1162/tacl_a_00023",
    pages = "317--328",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/natural_questions

  • Deskripsi konfigurasi : Korpus NQ berisi pertanyaan dari pengguna sebenarnya, dan memerlukan sistem QA untuk membaca dan memahami seluruh artikel Wikipedia yang mungkin berisi atau tidak berisi jawaban atas pertanyaan tersebut. Dimasukkannya pertanyaan pengguna nyata, dan persyaratan bahwa solusi harus membaca seluruh halaman untuk menemukan jawabannya, menyebabkan NQ menjadi tugas yang lebih realistis dan menantang daripada kumpulan data QA sebelumnya.

  • Ukuran unduhan : 6.95 MiB

  • Ukuran dataset : 9.88 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 96.075
'validation' 2.295
  • Kutipan :
@article{kwiatkowski-etal-2019-natural,
    title = "Natural Questions: A Benchmark for Question Answering Research",
    author = "Kwiatkowski, Tom  and
      Palomaki, Jennimaria  and
      Redfield, Olivia  and
      Collins, Michael  and
      Parikh, Ankur  and
      Alberti, Chris  and
      Epstein, Danielle  and
      Polosukhin, Illia  and
      Devlin, Jacob  and
      Lee, Kenton  and
      Toutanova, Kristina  and
      Jones, Llion  and
      Kelcey, Matthew  and
      Chang, Ming-Wei  and
      Dai, Andrew M.  and
      Uszkoreit, Jakob  and
      Le, Quoc  and
      Petrov, Slav",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "7",
    year = "2019",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q19-1026",
    doi = "10.1162/tacl_a_00276",
    pages = "452--466",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/natural_questions_direct_ans

  • Deskripsi konfigurasi : Korpus NQ berisi pertanyaan dari pengguna sebenarnya, dan memerlukan sistem QA untuk membaca dan memahami seluruh artikel Wikipedia yang mungkin berisi atau tidak berisi jawaban atas pertanyaan tersebut. Dimasukkannya pertanyaan pengguna nyata, dan persyaratan bahwa solusi harus membaca seluruh halaman untuk menemukan jawabannya, menyebabkan NQ menjadi tugas yang lebih realistis dan menantang daripada kumpulan data QA sebelumnya. Versi ini terdiri dari pertanyaan jawaban langsung.

  • Ukuran unduhan : 6.82 MiB

  • Ukuran dataset : 10.19 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 6.468
'train' 96.676
'validation' 10.693
  • Kutipan :
@article{kwiatkowski-etal-2019-natural,
    title = "Natural Questions: A Benchmark for Question Answering Research",
    author = "Kwiatkowski, Tom  and
      Palomaki, Jennimaria  and
      Redfield, Olivia  and
      Collins, Michael  and
      Parikh, Ankur  and
      Alberti, Chris  and
      Epstein, Danielle  and
      Polosukhin, Illia  and
      Devlin, Jacob  and
      Lee, Kenton  and
      Toutanova, Kristina  and
      Jones, Llion  and
      Kelcey, Matthew  and
      Chang, Ming-Wei  and
      Dai, Andrew M.  and
      Uszkoreit, Jakob  and
      Le, Quoc  and
      Petrov, Slav",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "7",
    year = "2019",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q19-1026",
    doi = "10.1162/tacl_a_00276",
    pages = "452--466",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/natural_questions_direct_ans_test

  • Deskripsi konfigurasi : Korpus NQ berisi pertanyaan dari pengguna sebenarnya, dan memerlukan sistem QA untuk membaca dan memahami seluruh artikel Wikipedia yang mungkin berisi atau tidak berisi jawaban atas pertanyaan tersebut. Dimasukkannya pertanyaan pengguna nyata, dan persyaratan bahwa solusi harus membaca seluruh halaman untuk menemukan jawabannya, menyebabkan NQ menjadi tugas yang lebih realistis dan menantang daripada kumpulan data QA sebelumnya. Versi ini terdiri dari pertanyaan jawaban langsung.

  • Ukuran unduhan : 6.82 MiB

  • Ukuran dataset : 10.19 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 6.468
'train' 96.676
'validation' 10.693
  • Kutipan :
@article{kwiatkowski-etal-2019-natural,
    title = "Natural Questions: A Benchmark for Question Answering Research",
    author = "Kwiatkowski, Tom  and
      Palomaki, Jennimaria  and
      Redfield, Olivia  and
      Collins, Michael  and
      Parikh, Ankur  and
      Alberti, Chris  and
      Epstein, Danielle  and
      Polosukhin, Illia  and
      Devlin, Jacob  and
      Lee, Kenton  and
      Toutanova, Kristina  and
      Jones, Llion  and
      Kelcey, Matthew  and
      Chang, Ming-Wei  and
      Dai, Andrew M.  and
      Uszkoreit, Jakob  and
      Le, Quoc  and
      Petrov, Slav",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "7",
    year = "2019",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q19-1026",
    doi = "10.1162/tacl_a_00276",
    pages = "452--466",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/natural_questions_with_dpr_para

  • Deskripsi konfigurasi : Korpus NQ berisi pertanyaan dari pengguna sebenarnya, dan memerlukan sistem QA untuk membaca dan memahami seluruh artikel Wikipedia yang mungkin berisi atau tidak berisi jawaban atas pertanyaan tersebut. Dimasukkannya pertanyaan pengguna nyata, dan persyaratan bahwa solusi harus membaca seluruh halaman untuk menemukan jawabannya, menyebabkan NQ menjadi tugas yang lebih realistis dan menantang daripada kumpulan data QA sebelumnya. Versi ini menyertakan paragraf tambahan (diperoleh dengan menggunakan mesin pencarian DPR) untuk menambah setiap pertanyaan.

  • Ukuran unduhan : 319.22 MiB

  • Ukuran dataset : 322.91 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'train' 96.676
'validation' 10.693
  • Kutipan :
@article{kwiatkowski-etal-2019-natural,
    title = "Natural Questions: A Benchmark for Question Answering Research",
    author = "Kwiatkowski, Tom  and
      Palomaki, Jennimaria  and
      Redfield, Olivia  and
      Collins, Michael  and
      Parikh, Ankur  and
      Alberti, Chris  and
      Epstein, Danielle  and
      Polosukhin, Illia  and
      Devlin, Jacob  and
      Lee, Kenton  and
      Toutanova, Kristina  and
      Jones, Llion  and
      Kelcey, Matthew  and
      Chang, Ming-Wei  and
      Dai, Andrew M.  and
      Uszkoreit, Jakob  and
      Le, Quoc  and
      Petrov, Slav",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "7",
    year = "2019",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q19-1026",
    doi = "10.1162/tacl_a_00276",
    pages = "452--466",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/natural_questions_with_dpr_para_test

  • Deskripsi konfigurasi : Korpus NQ berisi pertanyaan dari pengguna sebenarnya, dan memerlukan sistem QA untuk membaca dan memahami seluruh artikel Wikipedia yang mungkin berisi atau tidak berisi jawaban atas pertanyaan tersebut. Dimasukkannya pertanyaan pengguna nyata, dan persyaratan bahwa solusi harus membaca seluruh halaman untuk menemukan jawabannya, menyebabkan NQ menjadi tugas yang lebih realistis dan menantang daripada kumpulan data QA sebelumnya. Versi ini menyertakan paragraf tambahan (diperoleh dengan menggunakan mesin pencarian DPR) untuk menambah setiap pertanyaan.

  • Ukuran unduhan : 306.94 MiB

  • Ukuran dataset : 310.48 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'test' 6.468
'train' 96.676
  • Kutipan :
@article{kwiatkowski-etal-2019-natural,
    title = "Natural Questions: A Benchmark for Question Answering Research",
    author = "Kwiatkowski, Tom  and
      Palomaki, Jennimaria  and
      Redfield, Olivia  and
      Collins, Michael  and
      Parikh, Ankur  and
      Alberti, Chris  and
      Epstein, Danielle  and
      Polosukhin, Illia  and
      Devlin, Jacob  and
      Lee, Kenton  and
      Toutanova, Kristina  and
      Jones, Llion  and
      Kelcey, Matthew  and
      Chang, Ming-Wei  and
      Dai, Andrew M.  and
      Uszkoreit, Jakob  and
      Le, Quoc  and
      Petrov, Slav",
    journal = "Transactions of the Association for Computational Linguistics",
    volume = "7",
    year = "2019",
    address = "Cambridge, MA",
    publisher = "MIT Press",
    url = "https://aclanthology.org/Q19-1026",
    doi = "10.1162/tacl_a_00276",
    pages = "452--466",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/newsqa

  • Deskripsi konfigurasi : NewsQA adalah kumpulan data pemahaman mesin yang menantang dari pasangan pertanyaan-jawaban yang dihasilkan manusia. Crowdworker memberikan pertanyaan dan jawaban berdasarkan serangkaian artikel berita dari CNN, dengan jawaban yang terdiri dari rangkaian teks dari artikel terkait.

  • Ukuran unduhan : 283.33 MiB

  • Ukuran dataset : 285.94 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'train' 75.882
'validation' 4.309
  • Kutipan :
@inproceedings{trischler-etal-2017-newsqa,
    title = "{N}ews{QA}: A Machine Comprehension Dataset",
    author = "Trischler, Adam  and
      Wang, Tong  and
      Yuan, Xingdi  and
      Harris, Justin  and
      Sordoni, Alessandro  and
      Bachman, Philip  and
      Suleman, Kaheer",
    booktitle = "Proceedings of the 2nd Workshop on Representation Learning for {NLP}",
    month = aug,
    year = "2017",
    address = "Vancouver, Canada",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/W17-2623",
    doi = "10.18653/v1/W17-2623",
    pages = "191--200",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/openbookqa

  • Deskripsi Config : OpenBookQA bertujuan untuk mempromosikan penelitian dalam menjawab pertanyaan tingkat lanjut, menyelidiki pemahaman yang lebih dalam tentang topik (dengan fakta-fakta penting yang dirangkum sebagai buku terbuka, juga dilengkapi dengan kumpulan data) dan bahasa yang digunakan untuk menyatakannya. Secara khusus, ini berisi pertanyaan yang membutuhkan penalaran multi-langkah, penggunaan pengetahuan umum dan akal sehat tambahan, dan pemahaman teks yang kaya. OpenBookQA adalah jenis baru kumpulan data penjawab pertanyaan yang dimodelkan setelah ujian buku terbuka untuk menilai pemahaman manusia tentang suatu subjek.

  • Ukuran unduhan : 942.34 KiB

  • Ukuran dataset : 1.11 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 500
'train' 4.957
'validation' 500
  • Kutipan :
@inproceedings{mihaylov-etal-2018-suit,
    title = "Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering",
    author = "Mihaylov, Todor  and
      Clark, Peter  and
      Khot, Tushar  and
      Sabharwal, Ashish",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D18-1260",
    doi = "10.18653/v1/D18-1260",
    pages = "2381--2391",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/openbookqa_dev

  • Deskripsi Config : OpenBookQA bertujuan untuk mempromosikan penelitian dalam menjawab pertanyaan tingkat lanjut, menyelidiki pemahaman yang lebih dalam tentang topik (dengan fakta-fakta penting yang dirangkum sebagai buku terbuka, juga dilengkapi dengan kumpulan data) dan bahasa yang digunakan untuk menyatakannya. Secara khusus, ini berisi pertanyaan yang membutuhkan penalaran multi-langkah, penggunaan pengetahuan umum dan akal sehat tambahan, dan pemahaman teks yang kaya. OpenBookQA adalah jenis baru kumpulan data penjawab pertanyaan yang dimodelkan setelah ujian buku terbuka untuk menilai pemahaman manusia tentang suatu subjek.

  • Ukuran unduhan : 942.34 KiB

  • Ukuran dataset : 1.11 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 500
'train' 4.957
'validation' 500
  • Kutipan :
@inproceedings{mihaylov-etal-2018-suit,
    title = "Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering",
    author = "Mihaylov, Todor  and
      Clark, Peter  and
      Khot, Tushar  and
      Sabharwal, Ashish",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D18-1260",
    doi = "10.18653/v1/D18-1260",
    pages = "2381--2391",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/openbookqa_with_ir

  • Deskripsi Config : OpenBookQA bertujuan untuk mempromosikan penelitian dalam menjawab pertanyaan tingkat lanjut, menyelidiki pemahaman yang lebih dalam tentang topik (dengan fakta-fakta penting yang dirangkum sebagai buku terbuka, juga dilengkapi dengan kumpulan data) dan bahasa yang digunakan untuk menyatakannya. Secara khusus, ini berisi pertanyaan yang membutuhkan penalaran multi-langkah, penggunaan pengetahuan umum dan akal sehat tambahan, dan pemahaman teks yang kaya. OpenBookQA adalah jenis baru kumpulan data penjawab pertanyaan yang dimodelkan setelah ujian buku terbuka untuk menilai pemahaman manusia tentang suatu subjek. Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 6.08 MiB

  • Ukuran dataset : 6.28 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 500
'train' 4.957
'validation' 500
  • Kutipan :
@inproceedings{mihaylov-etal-2018-suit,
    title = "Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering",
    author = "Mihaylov, Todor  and
      Clark, Peter  and
      Khot, Tushar  and
      Sabharwal, Ashish",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D18-1260",
    doi = "10.18653/v1/D18-1260",
    pages = "2381--2391",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/openbookqa_with_ir_dev

  • Deskripsi Config : OpenBookQA bertujuan untuk mempromosikan penelitian dalam menjawab pertanyaan tingkat lanjut, menyelidiki pemahaman yang lebih dalam tentang topik (dengan fakta-fakta penting yang dirangkum sebagai buku terbuka, juga dilengkapi dengan kumpulan data) dan bahasa yang digunakan untuk menyatakannya. Secara khusus, ini berisi pertanyaan yang membutuhkan penalaran multi-langkah, penggunaan pengetahuan umum dan akal sehat tambahan, dan pemahaman teks yang kaya. OpenBookQA adalah jenis baru kumpulan data penjawab pertanyaan yang dimodelkan setelah ujian buku terbuka untuk menilai pemahaman manusia tentang suatu subjek. Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 6.08 MiB

  • Ukuran dataset : 6.28 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 500
'train' 4.957
'validation' 500
  • Kutipan :
@inproceedings{mihaylov-etal-2018-suit,
    title = "Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering",
    author = "Mihaylov, Todor  and
      Clark, Peter  and
      Khot, Tushar  and
      Sabharwal, Ashish",
    booktitle = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
    month = oct # "-" # nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D18-1260",
    doi = "10.18653/v1/D18-1260",
    pages = "2381--2391",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/physical_iqa

  • Deskripsi konfigurasi : Ini adalah kumpulan data untuk kemajuan pembandingan dalam pemahaman akal sehat fisik. Tugas yang mendasarinya adalah menjawab pertanyaan pilihan ganda: diberi pertanyaan q dan dua kemungkinan solusi s1, s2, model atau manusia harus memilih solusi yang paling tepat, yang salah satunya benar. Kumpulan data berfokus pada situasi sehari-hari dengan preferensi untuk solusi atipikal. Kumpulan data ini terinspirasi oleh instructables.com, yang memberikan petunjuk kepada pengguna tentang cara membuat, membuat, memanggang, atau memanipulasi objek menggunakan bahan sehari-hari. Anotator diminta untuk memberikan gangguan semantik atau pendekatan alternatif yang secara sintaksis dan topik serupa untuk memastikan pengetahuan fisik ditargetkan. Dataset selanjutnya dibersihkan dari artefak dasar menggunakan algoritma AFLite.

  • Ukuran unduhan : 6.01 MiB

  • Ukuran dataset : 6.59 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 16.113
'validation' 1.838
  • Kutipan :
@inproceedings{bisk2020piqa,
    title={Piqa: Reasoning about physical commonsense in natural language},
    author={Bisk, Yonatan and Zellers, Rowan and Gao, Jianfeng and Choi, Yejin and others},
    booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
    volume={34},
    number={05},
    pages={7432--7439},
    year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/qasc

  • Deskripsi konfigurasi : QASC adalah kumpulan data penjawab pertanyaan dengan fokus pada komposisi kalimat. Ini terdiri dari pertanyaan pilihan ganda 8 arah tentang sains sekolah dasar, dan dilengkapi dengan kumpulan 17 juta kalimat.

  • Ukuran unduhan : 1.75 MiB

  • Ukuran dataset : 2.09 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 920
'train' 8.134
'validation' 926
  • Kutipan :
@inproceedings{khot2020qasc,
    title={Qasc: A dataset for question answering via sentence composition},
    author={Khot, Tushar and Clark, Peter and Guerquin, Michal and Jansen, Peter and Sabharwal, Ashish},
    booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
    volume={34},
    number={05},
    pages={8082--8090},
    year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/qasc_test

  • Deskripsi konfigurasi : QASC adalah kumpulan data penjawab pertanyaan dengan fokus pada komposisi kalimat. Ini terdiri dari pertanyaan pilihan ganda 8 arah tentang sains sekolah dasar, dan dilengkapi dengan kumpulan 17 juta kalimat.

  • Ukuran unduhan : 1.75 MiB

  • Ukuran dataset : 2.09 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 920
'train' 8.134
'validation' 926
  • Kutipan :
@inproceedings{khot2020qasc,
    title={Qasc: A dataset for question answering via sentence composition},
    author={Khot, Tushar and Clark, Peter and Guerquin, Michal and Jansen, Peter and Sabharwal, Ashish},
    booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
    volume={34},
    number={05},
    pages={8082--8090},
    year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/qasc_with_ir

  • Deskripsi konfigurasi : QASC adalah kumpulan data penjawab pertanyaan dengan fokus pada komposisi kalimat. Ini terdiri dari pertanyaan pilihan ganda 8 arah tentang sains sekolah dasar, dan dilengkapi dengan kumpulan 17 juta kalimat. Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 16.95 MiB

  • Ukuran dataset : 17.30 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 920
'train' 8.134
'validation' 926
  • Kutipan :
@inproceedings{khot2020qasc,
    title={Qasc: A dataset for question answering via sentence composition},
    author={Khot, Tushar and Clark, Peter and Guerquin, Michal and Jansen, Peter and Sabharwal, Ashish},
    booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
    volume={34},
    number={05},
    pages={8082--8090},
    year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/qasc_with_ir_test

  • Deskripsi konfigurasi : QASC adalah kumpulan data penjawab pertanyaan dengan fokus pada komposisi kalimat. Ini terdiri dari pertanyaan pilihan ganda 8 arah tentang sains sekolah dasar, dan dilengkapi dengan kumpulan 17 juta kalimat. Versi ini menyertakan paragraf yang diambil melalui sistem pencarian informasi sebagai bukti tambahan.

  • Ukuran unduhan : 16.95 MiB

  • Ukuran dataset : 17.30 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 920
'train' 8.134
'validation' 926
  • Kutipan :
@inproceedings{khot2020qasc,
    title={Qasc: A dataset for question answering via sentence composition},
    author={Khot, Tushar and Clark, Peter and Guerquin, Michal and Jansen, Peter and Sabharwal, Ashish},
    booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
    volume={34},
    number={05},
    pages={8082--8090},
    year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/quoref

  • Deskripsi konfigurasi : Kumpulan data ini menguji kemampuan penalaran inti dari sistem pemahaman bacaan. Dalam tolok ukur pemilihan rentang ini yang berisi pertanyaan tentang paragraf dari Wikipedia, sistem harus menyelesaikan referensi keras sebelum memilih rentang yang sesuai dalam paragraf untuk menjawab pertanyaan.

  • Ukuran unduhan : 51.43 MiB

  • Ukuran dataset : 52.29 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 22.265
'validation' 2.768
  • Kutipan :
@inproceedings{dasigi-etal-2019-quoref,
    title = "{Q}uoref: A Reading Comprehension Dataset with Questions Requiring Coreferential Reasoning",
    author = "Dasigi, Pradeep  and
      Liu, Nelson F.  and
      Marasovi{'c}, Ana  and
      Smith, Noah A.  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-1606",
    doi = "10.18653/v1/D19-1606",
    pages = "5925--5932",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/race_string

  • Deskripsi konfigurasi : Ras adalah kumpulan data pemahaman bacaan berskala besar. Kumpulan data dikumpulkan dari ujian bahasa Inggris di China, yang dirancang untuk siswa sekolah menengah dan sekolah menengah atas. Dataset dapat berfungsi sebagai set pelatihan dan pengujian untuk pemahaman mesin.

  • Ukuran unduhan : 167.97 MiB

  • Ukuran dataset : 171.23 MiB

  • Auto-cached ( dokumentasi ): Ya (test, validasi), Hanya ketika shuffle_files=False (train)

  • Perpecahan :

Membelah Contoh
'test' 4.934
'train' 87.863
'validation' 4.887
  • Kutipan :
@inproceedings{lai-etal-2017-race,
    title = "{RACE}: Large-scale {R}e{A}ding Comprehension Dataset From Examinations",
    author = "Lai, Guokun  and
      Xie, Qizhe  and
      Liu, Hanxiao  and
      Yang, Yiming  and
      Hovy, Eduard",
    booktitle = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
    month = sep,
    year = "2017",
    address = "Copenhagen, Denmark",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D17-1082",
    doi = "10.18653/v1/D17-1082",
    pages = "785--794",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/race_string_dev

  • Deskripsi konfigurasi : Ras adalah kumpulan data pemahaman bacaan berskala besar. Kumpulan data dikumpulkan dari ujian bahasa Inggris di China, yang dirancang untuk siswa sekolah menengah dan sekolah menengah atas. Dataset dapat berfungsi sebagai set pelatihan dan pengujian untuk pemahaman mesin.

  • Ukuran unduhan : 167.97 MiB

  • Ukuran dataset : 171.23 MiB

  • Auto-cached ( dokumentasi ): Ya (test, validasi), Hanya ketika shuffle_files=False (train)

  • Perpecahan :

Membelah Contoh
'test' 4.934
'train' 87.863
'validation' 4.887
  • Kutipan :
@inproceedings{lai-etal-2017-race,
    title = "{RACE}: Large-scale {R}e{A}ding Comprehension Dataset From Examinations",
    author = "Lai, Guokun  and
      Xie, Qizhe  and
      Liu, Hanxiao  and
      Yang, Yiming  and
      Hovy, Eduard",
    booktitle = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
    month = sep,
    year = "2017",
    address = "Copenhagen, Denmark",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D17-1082",
    doi = "10.18653/v1/D17-1082",
    pages = "785--794",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/ropes

  • Deskripsi konfigurasi : Kumpulan data ini menguji kemampuan sistem untuk menerapkan pengetahuan dari bagian teks ke situasi baru. Suatu sistem disajikan dengan bagian latar belakang yang berisi hubungan kausal atau kualitatif (misalnya, "penyerbuk hewan meningkatkan efisiensi pembuahan pada bunga"), situasi baru yang menggunakan latar belakang ini, dan pertanyaan yang memerlukan penalaran tentang efek hubungan dalam bagian latar belakang dalam konteks situasi.

  • Ukuran unduhan : 12.91 MiB

  • Ukuran dataset : 13.35 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 10.924
'validation' 1.688
  • Kutipan :
@inproceedings{lin-etal-2019-reasoning,
    title = "Reasoning Over Paragraph Effects in Situations",
    author = "Lin, Kevin  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Gardner, Matt",
    booktitle = "Proceedings of the 2nd Workshop on Machine Reading for Question Answering",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-5808",
    doi = "10.18653/v1/D19-5808",
    pages = "58--62",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/social_iqa

  • Deskripsi konfigurasi : Ini adalah tolok ukur skala besar untuk penalaran akal sehat tentang situasi sosial. Social IQa berisi soal pilihan ganda untuk menggali kecerdasan emosional dan sosial dalam berbagai situasi sehari-hari. Melalui crowdsourcing, pertanyaan akal sehat bersama dengan jawaban benar dan salah tentang interaksi sosial dikumpulkan, menggunakan kerangka kerja baru yang mengurangi artefak gaya dalam jawaban yang salah dengan meminta pekerja untuk memberikan jawaban yang benar untuk pertanyaan yang berbeda namun terkait.

  • Ukuran unduhan : 7.08 MiB

  • Ukuran dataset : 8.22 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 33.410
'validation' 1.954
  • Kutipan :
@inproceedings{sap-etal-2019-social,
    title = "Social {IQ}a: Commonsense Reasoning about Social Interactions",
    author = "Sap, Maarten  and
      Rashkin, Hannah  and
      Chen, Derek  and
      Le Bras, Ronan  and
      Choi, Yejin",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D19-1454",
    doi = "10.18653/v1/D19-1454",
    pages = "4463--4473",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/squad1_1

  • Deskripsi konfigurasi : Ini adalah kumpulan data pemahaman bacaan yang terdiri dari pertanyaan yang diajukan oleh crowdworker pada serangkaian artikel Wikipedia, di mana jawaban untuk setiap pertanyaan adalah segmen teks dari bagian bacaan yang sesuai.

  • Ukuran unduhan : 80.62 MiB

  • Ukuran dataset : 83.99 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 87.514
'validation' 10.570
  • Kutipan :
@inproceedings{rajpurkar-etal-2016-squad,
    title = "{SQ}u{AD}: 100,000+ Questions for Machine Comprehension of Text",
    author = "Rajpurkar, Pranav  and
      Zhang, Jian  and
      Lopyrev, Konstantin  and
      Liang, Percy",
    booktitle = "Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing",
    month = nov,
    year = "2016",
    address = "Austin, Texas",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/D16-1264",
    doi = "10.18653/v1/D16-1264",
    pages = "2383--2392",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/squad2

  • Deskripsi konfigurasi : Kumpulan data ini menggabungkan kumpulan data Stanford Question Answering Dataset (SQuAD) asli dengan pertanyaan yang tidak dapat dijawab yang ditulis secara berlawanan oleh crowdworker agar terlihat mirip dengan yang dapat dijawab.

  • Ukuran unduhan : 116.56 MiB

  • Ukuran dataset : 121.43 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 130.149
'validation' 11.873
  • Kutipan :
@inproceedings{rajpurkar-etal-2018-know,
    title = "Know What You Don{'}t Know: Unanswerable Questions for {SQ}u{AD}",
    author = "Rajpurkar, Pranav  and
      Jia, Robin  and
      Liang, Percy",
    booktitle = "Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
    month = jul,
    year = "2018",
    address = "Melbourne, Australia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/P18-2124",
    doi = "10.18653/v1/P18-2124",
    pages = "784--789",
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/winogrande_l

  • Deskripsi Config : Dataset ini terinspirasi oleh desain Winograd Schema Challenge asli, tetapi disesuaikan untuk meningkatkan skala dan kekerasan dataset. Langkah-langkah kunci dari konstruksi dataset terdiri dari (1) prosedur crowdsourcing yang dirancang dengan hati-hati, diikuti oleh (2) pengurangan bias sistematis menggunakan algoritme AfLite baru yang menggeneralisasi asosiasi kata yang dapat dideteksi manusia menjadi asosiasi penyematan yang dapat dideteksi mesin. Set pelatihan dengan ukuran berbeda disediakan. Set ini sesuai dengan ukuran l .

  • Ukuran unduhan : 1.49 MiB

  • Ukuran dataset : 1.83 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 10.234
'validation' 1.267
  • Kutipan :
@inproceedings{sakaguchi2020winogrande,
  title={Winogrande: An adversarial winograd schema challenge at scale},
  author={Sakaguchi, Keisuke and Le Bras, Ronan and Bhagavatula, Chandra and Choi, Yejin},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={34},
  number={05},
  pages={8732--8740},
  year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/winogrande_m

  • Deskripsi Config : Dataset ini terinspirasi oleh desain Winograd Schema Challenge asli, tetapi disesuaikan untuk meningkatkan skala dan kekerasan dataset. Langkah-langkah kunci dari konstruksi dataset terdiri dari (1) prosedur crowdsourcing yang dirancang dengan hati-hati, diikuti oleh (2) pengurangan bias sistematis menggunakan algoritme AfLite baru yang menggeneralisasi asosiasi kata yang dapat dideteksi manusia menjadi asosiasi penyematan yang dapat dideteksi mesin. Set pelatihan dengan ukuran berbeda disediakan. Set ini sesuai dengan ukuran m .

  • Ukuran unduhan : 507.46 KiB

  • Ukuran dataset : 623.15 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'train' 2.558
'validation' 1.267
  • Kutipan :
@inproceedings{sakaguchi2020winogrande,
  title={Winogrande: An adversarial winograd schema challenge at scale},
  author={Sakaguchi, Keisuke and Le Bras, Ronan and Bhagavatula, Chandra and Choi, Yejin},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={34},
  number={05},
  pages={8732--8740},
  year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."

unified_qa/winogrande_s

  • Deskripsi Config : Dataset ini terinspirasi oleh desain Winograd Schema Challenge asli, tetapi disesuaikan untuk meningkatkan skala dan kekerasan dataset. Langkah-langkah kunci dari konstruksi dataset terdiri dari (1) prosedur crowdsourcing yang dirancang dengan hati-hati, diikuti oleh (2) pengurangan bias sistematis menggunakan algoritme AfLite baru yang menggeneralisasi asosiasi kata yang dapat dideteksi manusia menjadi asosiasi penyematan yang dapat dideteksi mesin. Set pelatihan dengan ukuran berbeda disediakan. Set ini sesuai dengan ukuran s .

  • Ukuran unduhan : 479.24 KiB

  • Ukuran dataset : 590.47 KiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 1.767
'train' 640
'validation' 1.267
  • Kutipan :
@inproceedings{sakaguchi2020winogrande,
  title={Winogrande: An adversarial winograd schema challenge at scale},
  author={Sakaguchi, Keisuke and Le Bras, Ronan and Bhagavatula, Chandra and Choi, Yejin},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={34},
  number={05},
  pages={8732--8740},
  year={2020}
}

@inproceedings{khashabi-etal-2020-unifiedqa,
    title = "{UNIFIEDQA}: Crossing Format Boundaries with a Single {QA} System",
    author = "Khashabi, Daniel  and
      Min, Sewon  and
      Khot, Tushar  and
      Sabharwal, Ashish  and
      Tafjord, Oyvind  and
      Clark, Peter  and
      Hajishirzi, Hannaneh",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.171",
    doi = "10.18653/v1/2020.findings-emnlp.171",
    pages = "1896--1907",
}

Note that each UnifiedQA dataset has its own citation. Please see the source to
see the correct citation for each contained dataset."