albertvillanova HF staff commited on
Commit
7063139
1 Parent(s): ed9bab1

Fix license/citation information of squadshifts dataset card (#5054)

Browse files

* Update license information in squadshifts dataset

* Update citation information in squadshifts dataset

* Update metadata JSON

Commit from https://github.com/huggingface/datasets/commit/61258078bcb80113f877e84745ed85ec85bb11d2

Files changed (3) hide show
  1. README.md +13 -7
  2. dataset_infos.json +1 -1
  3. squadshifts.py +20 -12
README.md CHANGED
@@ -7,7 +7,7 @@ language_creators:
7
  - crowdsourced
8
  - found
9
  license:
10
- - unknown
11
  multilinguality:
12
  - monolingual
13
  pretty_name: SQuAD-shifts
@@ -256,21 +256,27 @@ The data fields are the same among all splits.
256
 
257
  ### Licensing Information
258
 
259
- [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
260
 
261
  ### Citation Information
262
 
263
  ```
264
- @inproceedings{miller2020effect,
265
- author = {J. Miller and K. Krauth and B. Recht and L. Schmidt},
266
- booktitle = {International Conference on Machine Learning (ICML)},
267
  title = {The Effect of Natural Distribution Shift on Question Answering Models},
 
 
 
268
  year = {2020},
 
 
 
 
 
 
 
269
  }
270
-
271
  ```
272
 
273
-
274
  ### Contributions
275
 
276
  Thanks to [@thomwolf](https://github.com/thomwolf), [@lewtun](https://github.com/lewtun), [@millerjohnp](https://github.com/millerjohnp), [@albertvillanova](https://github.com/albertvillanova) for adding this dataset.
7
  - crowdsourced
8
  - found
9
  license:
10
+ - cc-by-4.0
11
  multilinguality:
12
  - monolingual
13
  pretty_name: SQuAD-shifts
256
 
257
  ### Licensing Information
258
 
259
+ All the datasets are distributed under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/legalcode) license.
260
 
261
  ### Citation Information
262
 
263
  ```
264
+ @InProceedings{pmlr-v119-miller20a,
 
 
265
  title = {The Effect of Natural Distribution Shift on Question Answering Models},
266
+ author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},
267
+ booktitle = {Proceedings of the 37th International Conference on Machine Learning},
268
+ pages = {6905--6916},
269
  year = {2020},
270
+ editor = {III, Hal Daumé and Singh, Aarti},
271
+ volume = {119},
272
+ series = {Proceedings of Machine Learning Research},
273
+ month = {13--18 Jul},
274
+ publisher = {PMLR},
275
+ pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},
276
+ url = {https://proceedings.mlr.press/v119/miller20a.html},
277
  }
 
278
  ```
279
 
 
280
  ### Contributions
281
 
282
  Thanks to [@thomwolf](https://github.com/thomwolf), [@lewtun](https://github.com/lewtun), [@millerjohnp](https://github.com/millerjohnp), [@albertvillanova](https://github.com/albertvillanova) for adding this dataset.
dataset_infos.json CHANGED
@@ -1 +1 @@
1
- {"new_wiki": {"description": "SquadShifts consists of four new test sets for the Stanford Question Answering Dataset (SQuAD) from four different domains: Wikipedia articles, New York \\ \nTimes articles, Reddit comments, and Amazon product reviews. Each dataset was generated using the same data generating pipeline, Amazon Mechanical Turk interface, and data cleaning code as the original SQuAD v1.1 dataset. The \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 dataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure robustness to natural distribution shifts. We encourage SQuAD model developers to also evaluate their methods on these new datasets! ", "citation": "@inproceedings{miller2020effect,\n author = {J. Miller and K. Krauth and B. Recht and L. Schmidt},\n booktitle = {International Conference on Machine Learning (ICML)},\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n year = {2020},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "supervised_keys": null, "builder_name": "squad_shifts", "config_name": "new_wiki", "version": {"version_str": "1.0.0", "description": "", "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 7861360, "num_examples": 7938, "dataset_name": "squad_shifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "dataset_size": 7861360, "size_in_bytes": 24366983}, "nyt": {"description": "SquadShifts consists of four new test sets for the Stanford Question Answering Dataset (SQuAD) from four different domains: Wikipedia articles, New York \\ \nTimes articles, Reddit comments, and Amazon product reviews. Each dataset was generated using the same data generating pipeline, Amazon Mechanical Turk interface, and data cleaning code as the original SQuAD v1.1 dataset. The \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 dataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure robustness to natural distribution shifts. We encourage SQuAD model developers to also evaluate their methods on these new datasets! ", "citation": "@inproceedings{miller2020effect,\n author = {J. Miller and K. Krauth and B. Recht and L. Schmidt},\n booktitle = {International Conference on Machine Learning (ICML)},\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n year = {2020},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "supervised_keys": null, "builder_name": "squad_shifts", "config_name": "nyt", "version": {"version_str": "1.0.0", "description": "", "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 10787650, "num_examples": 10065, "dataset_name": "squad_shifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "dataset_size": 10787650, "size_in_bytes": 27293273}, "reddit": {"description": "SquadShifts consists of four new test sets for the Stanford Question Answering Dataset (SQuAD) from four different domains: Wikipedia articles, New York \\ \nTimes articles, Reddit comments, and Amazon product reviews. Each dataset was generated using the same data generating pipeline, Amazon Mechanical Turk interface, and data cleaning code as the original SQuAD v1.1 dataset. The \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 dataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure robustness to natural distribution shifts. We encourage SQuAD model developers to also evaluate their methods on these new datasets! ", "citation": "@inproceedings{miller2020effect,\n author = {J. Miller and K. Krauth and B. Recht and L. Schmidt},\n booktitle = {International Conference on Machine Learning (ICML)},\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n year = {2020},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "supervised_keys": null, "builder_name": "squad_shifts", "config_name": "reddit", "version": {"version_str": "1.0.0", "description": "", "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 9469258, "num_examples": 9803, "dataset_name": "squad_shifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "dataset_size": 9469258, "size_in_bytes": 25974881}, "amazon": {"description": "SquadShifts consists of four new test sets for the Stanford Question Answering Dataset (SQuAD) from four different domains: Wikipedia articles, New York \\ \nTimes articles, Reddit comments, and Amazon product reviews. Each dataset was generated using the same data generating pipeline, Amazon Mechanical Turk interface, and data cleaning code as the original SQuAD v1.1 dataset. The \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 dataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure robustness to natural distribution shifts. We encourage SQuAD model developers to also evaluate their methods on these new datasets! ", "citation": "@inproceedings{miller2020effect,\n author = {J. Miller and K. Krauth and B. Recht and L. Schmidt},\n booktitle = {International Conference on Machine Learning (ICML)},\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n year = {2020},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "supervised_keys": null, "builder_name": "squad_shifts", "config_name": "amazon", "version": {"version_str": "1.0.0", "description": "", "datasets_version_to_prepare": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 9440353, "num_examples": 9885, "dataset_name": "squad_shifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "dataset_size": 9440353, "size_in_bytes": 25945976}}
1
+ {"new_wiki": {"description": "\\\nSquadShifts consists of four new test sets for the Stanford Question Answering \\\nDataset (SQuAD) from four different domains: Wikipedia articles, New York \\\nTimes articles, Reddit comments, and Amazon product reviews. Each dataset \\\nwas generated using the same data generating pipeline, Amazon Mechanical \\\nTurk interface, and data cleaning code as the original SQuAD v1.1 dataset. \\\nThe \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 \\\ndataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure \\\nrobustness to natural distribution shifts. We encourage SQuAD model developers \\\nto also evaluate their methods on these new datasets! \\\n", "citation": "@InProceedings{pmlr-v119-miller20a,\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},\n booktitle = {Proceedings of the 37th International Conference on Machine Learning},\n pages = {6905--6916},\n year = {2020},\n editor = {III, Hal Daum\u00e9 and Singh, Aarti},\n volume = {119},\n series = {Proceedings of Machine Learning Research},\n month = {13--18 Jul},\n publisher = {PMLR},\n pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},\n url = {https://proceedings.mlr.press/v119/miller20a.html},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "CC-BY-4.0", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "squadshifts", "config_name": "new_wiki", "version": {"version_str": "1.0.0", "description": "", "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 7865203, "num_examples": 7938, "dataset_name": "squadshifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "post_processing_size": null, "dataset_size": 7865203, "size_in_bytes": 24370826}, "nyt": {"description": "\\\nSquadShifts consists of four new test sets for the Stanford Question Answering \\\nDataset (SQuAD) from four different domains: Wikipedia articles, New York \\\nTimes articles, Reddit comments, and Amazon product reviews. Each dataset \\\nwas generated using the same data generating pipeline, Amazon Mechanical \\\nTurk interface, and data cleaning code as the original SQuAD v1.1 dataset. \\\nThe \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 \\\ndataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure \\\nrobustness to natural distribution shifts. We encourage SQuAD model developers \\\nto also evaluate their methods on these new datasets! \\\n", "citation": "@InProceedings{pmlr-v119-miller20a,\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},\n booktitle = {Proceedings of the 37th International Conference on Machine Learning},\n pages = {6905--6916},\n year = {2020},\n editor = {III, Hal Daum\u00e9 and Singh, Aarti},\n volume = {119},\n series = {Proceedings of Machine Learning Research},\n month = {13--18 Jul},\n publisher = {PMLR},\n pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},\n url = {https://proceedings.mlr.press/v119/miller20a.html},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "CC-BY-4.0", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "squadshifts", "config_name": "nyt", "version": {"version_str": "1.0.0", "description": "", "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 10792550, "num_examples": 10065, "dataset_name": "squadshifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "post_processing_size": null, "dataset_size": 10792550, "size_in_bytes": 27298173}, "reddit": {"description": "\\\nSquadShifts consists of four new test sets for the Stanford Question Answering \\\nDataset (SQuAD) from four different domains: Wikipedia articles, New York \\\nTimes articles, Reddit comments, and Amazon product reviews. Each dataset \\\nwas generated using the same data generating pipeline, Amazon Mechanical \\\nTurk interface, and data cleaning code as the original SQuAD v1.1 dataset. \\\nThe \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 \\\ndataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure \\\nrobustness to natural distribution shifts. We encourage SQuAD model developers \\\nto also evaluate their methods on these new datasets! \\\n", "citation": "@InProceedings{pmlr-v119-miller20a,\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},\n booktitle = {Proceedings of the 37th International Conference on Machine Learning},\n pages = {6905--6916},\n year = {2020},\n editor = {III, Hal Daum\u00e9 and Singh, Aarti},\n volume = {119},\n series = {Proceedings of Machine Learning Research},\n month = {13--18 Jul},\n publisher = {PMLR},\n pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},\n url = {https://proceedings.mlr.press/v119/miller20a.html},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "CC-BY-4.0", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "squadshifts", "config_name": "reddit", "version": {"version_str": "1.0.0", "description": "", "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 9473946, "num_examples": 9803, "dataset_name": "squadshifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "post_processing_size": null, "dataset_size": 9473946, "size_in_bytes": 25979569}, "amazon": {"description": "\\\nSquadShifts consists of four new test sets for the Stanford Question Answering \\\nDataset (SQuAD) from four different domains: Wikipedia articles, New York \\\nTimes articles, Reddit comments, and Amazon product reviews. Each dataset \\\nwas generated using the same data generating pipeline, Amazon Mechanical \\\nTurk interface, and data cleaning code as the original SQuAD v1.1 dataset. \\\nThe \"new-wikipedia\" dataset measures overfitting on the original SQuAD v1.1 \\\ndataset. The \"new-york-times\", \"reddit\", and \"amazon\" datasets measure \\\nrobustness to natural distribution shifts. We encourage SQuAD model developers \\\nto also evaluate their methods on these new datasets! \\\n", "citation": "@InProceedings{pmlr-v119-miller20a,\n title = {The Effect of Natural Distribution Shift on Question Answering Models},\n author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},\n booktitle = {Proceedings of the 37th International Conference on Machine Learning},\n pages = {6905--6916},\n year = {2020},\n editor = {III, Hal Daum\u00e9 and Singh, Aarti},\n volume = {119},\n series = {Proceedings of Machine Learning Research},\n month = {13--18 Jul},\n publisher = {PMLR},\n pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},\n url = {https://proceedings.mlr.press/v119/miller20a.html},\n}\n", "homepage": "https://modestyachts.github.io/squadshifts-website/index.html", "license": "CC-BY-4.0", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "squadshifts", "config_name": "amazon", "version": {"version_str": "1.0.0", "description": "", "major": 1, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 9445004, "num_examples": 9885, "dataset_name": "squadshifts"}}, "download_checksums": {"https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/new_wiki_v1.0.json": {"num_bytes": 3607393, "checksum": "1d50a0ca4cdab88d038b52ea8a2d0510af6038d7f41a0a256b9659237769d0ee"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/nyt_v1.0.json": {"num_bytes": 4550058, "checksum": "fd55ae252cc735a695fab4473d4cd605b154183ac3b3eb63068f53cfc86f36d6"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/reddit_v1.0.json": {"num_bytes": 4170330, "checksum": "b4ac346dbf1816fdec80fdcf91d06a4b754fa51e47960fa8614a94f37ea42a4b"}, "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/amazon_reviews_v1.0.json": {"num_bytes": 4177842, "checksum": "9db7bdee84a91f0fb80b9e74f365042fbbb217cb8e0a7e73bb071245472589b6"}}, "download_size": 16505623, "post_processing_size": null, "dataset_size": 9445004, "size_in_bytes": 25950627}}
squadshifts.py CHANGED
@@ -25,15 +25,6 @@ import datasets
25
  logger = datasets.logging.get_logger(__name__)
26
 
27
 
28
- _CITATION = """\
29
- @inproceedings{miller2020effect,
30
- author = {J. Miller and K. Krauth and B. Recht and L. Schmidt},
31
- booktitle = {International Conference on Machine Learning (ICML)},
32
- title = {The Effect of Natural Distribution Shift on Question Answering Models},
33
- year = {2020},
34
- }
35
- """
36
-
37
  _DESCRIPTION = r"""\
38
  SquadShifts consists of four new test sets for the Stanford Question Answering \
39
  Dataset (SQuAD) from four different domains: Wikipedia articles, New York \
@@ -46,6 +37,25 @@ robustness to natural distribution shifts. We encourage SQuAD model developers \
46
  to also evaluate their methods on these new datasets! \
47
  """
48
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
49
  _URL = "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/"
50
  _URLS = {
51
  "new_wiki": _URL + "new_wiki_v1.0.json",
@@ -110,10 +120,8 @@ class SquadShifts(datasets.GeneratorBasedBuilder):
110
  ),
111
  }
112
  ),
113
- # No default supervised_keys (as we have to pass both question
114
- # and context as input).
115
- supervised_keys=None,
116
  homepage="https://modestyachts.github.io/squadshifts-website/index.html",
 
117
  citation=_CITATION,
118
  )
119
 
25
  logger = datasets.logging.get_logger(__name__)
26
 
27
 
 
 
 
 
 
 
 
 
 
28
  _DESCRIPTION = r"""\
29
  SquadShifts consists of four new test sets for the Stanford Question Answering \
30
  Dataset (SQuAD) from four different domains: Wikipedia articles, New York \
37
  to also evaluate their methods on these new datasets! \
38
  """
39
 
40
+ _LICENSE = "CC-BY-4.0"
41
+
42
+ _CITATION = """\
43
+ @InProceedings{pmlr-v119-miller20a,
44
+ title = {The Effect of Natural Distribution Shift on Question Answering Models},
45
+ author = {Miller, John and Krauth, Karl and Recht, Benjamin and Schmidt, Ludwig},
46
+ booktitle = {Proceedings of the 37th International Conference on Machine Learning},
47
+ pages = {6905--6916},
48
+ year = {2020},
49
+ editor = {III, Hal Daumé and Singh, Aarti},
50
+ volume = {119},
51
+ series = {Proceedings of Machine Learning Research},
52
+ month = {13--18 Jul},
53
+ publisher = {PMLR},
54
+ pdf = {http://proceedings.mlr.press/v119/miller20a/miller20a.pdf},
55
+ url = {https://proceedings.mlr.press/v119/miller20a.html},
56
+ }
57
+ """
58
+
59
  _URL = "https://raw.githubusercontent.com/modestyachts/squadshifts-website/master/datasets/"
60
  _URLS = {
61
  "new_wiki": _URL + "new_wiki_v1.0.json",
120
  ),
121
  }
122
  ),
 
 
 
123
  homepage="https://modestyachts.github.io/squadshifts-website/index.html",
124
+ license=_LICENSE,
125
  citation=_CITATION,
126
  )
127