Add missing features to commonsense_qa dataset (#4280)

* Fix homepage URL

* Clean code

* Add missing features

* Update metadata

* Fix dataset card tags

* Update dummy data

* Fix style

Commit from https://github.com/huggingface/datasets/commit/bc55315c2ab708cbe295e990160cd2bb7eefaccc

README.md

@@ -1,13 +1,30 @@
 ---
+annotations_creators:
+- crowdsourced
+language_creators:
+- crowdsourced
 languages:
 - en
-paperswithcode_id: commonsenseqa
+licenses:
+- unknown
+multilinguality:
+- monolingual
 pretty_name: CommonsenseQA
+size_categories:
+- 1K<n<10K
+source_datasets:
+- original
+task_categories:
+- question-answering
+task_ids:
+- open-domain-qa
+paperswithcode_id: commonsenseqa
 ---
 
 # Dataset Card for "commonsense_qa"
 
 ## Table of Contents
+- [Table of Contents](#table-of-contents)
 - [Dataset Description](#dataset-description)
   - [Dataset Summary](#dataset-summary)
   - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
@@ -33,9 +50,9 @@ pretty_name: CommonsenseQA
 
 ## Dataset Description
 
-- **Homepage:** [https://www.tau-nlp.org/commonsenseqa](https://www.tau-nlp.org/commonsenseqa)
-- **Repository:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
-- **Paper:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+- **Homepage:** https://www.tau-nlp.org/commonsenseqa
+- **Repository:** https://github.com/jonathanherzig/commonsenseqa
+- **Paper:** https://arxiv.org/abs/1811.00937
 - **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
 - **Size of downloaded dataset files:** 4.46 MB
 - **Size of the generated dataset:** 2.08 MB
@@ -44,9 +61,9 @@ pretty_name: CommonsenseQA
 ### Dataset Summary
 
 CommonsenseQA is a new multiple-choice question answering dataset that requires different types of commonsense knowledge
- to predict the correct answers . It contains 12,102 questions with one correct answer and four distractor answers.
- The dataset is provided in two major training/validation/testing set splits: "Random split" which is the main evaluation
-  split, and "Question token split", see paper for details.
+to predict the correct answers . It contains 12,102 questions with one correct answer and four distractor answers.
+The dataset is provided in two major training/validation/testing set splits: "Random split" which is the main evaluation
+split, and "Question token split", see paper for details.
 
 ### Supported Tasks and Leaderboards
 
@@ -54,7 +71,7 @@ CommonsenseQA is a new multiple-choice question answering dataset that requires
 
 ### Languages
 
-[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+The dataset is in English (`en`).
 
 ## Dataset Structure
 
@@ -66,16 +83,14 @@ CommonsenseQA is a new multiple-choice question answering dataset that requires
 - **Size of the generated dataset:** 2.08 MB
 - **Total amount of disk used:** 6.54 MB
 
-An example of 'train' looks as follows.
+An example of 'train' looks as follows:
 ```
-{
-    "answerKey": "B",
-    "choices": {
-        "label": ["A", "B", "C", "D", "E"],
-        "text": ["mildred's coffee shop", "mexico", "diner", "kitchen", "canteen"]
-    },
-    "question": "In what Spanish speaking North American country can you get a great cup of coffee?"
-}
+{'id': '075e483d21c29a511267ef62bedc0461',
+ 'question': 'The sanctions against the school were a punishing blow, and they seemed to what the efforts the school had made to change?',
+ 'question_concept': 'punishing',
+ 'choices': {'label': ['A', 'B', 'C', 'D', 'E'],
+  'text': ['ignore', 'enforce', 'authoritarian', 'yell at', 'avoid']},
+ 'answerKey': 'A'}
 ```
 
 ### Data Fields
@@ -83,17 +98,19 @@ An example of 'train' looks as follows.
 The data fields are the same among all splits.
 
 #### default
-- `answerKey`: a `string` feature.
+- `id` (`str`): Unique ID.
 - `question`: a `string` feature.
+- `question_concept` (`str`): ConceptNet concept associated to the question.
 - `choices`: a dictionary feature containing:
   - `label`: a `string` feature.
   - `text`: a `string` feature.
+- `answerKey`: a `string` feature.
 
 ### Data Splits
 
-| name  |train|validation|test|
-|-------|----:|---------:|---:|
-|default| 9741|      1221|1140|
+| name    | train | validation | test |
+|---------|------:|-----------:|-----:|
+| default |  9741 |       1221 | 1140 |
 
 ## Dataset Creation
 
@@ -147,20 +164,31 @@ The data fields are the same among all splits.
 
 ### Licensing Information
 
-[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+Unknown.
 
 ### Citation Information
 
 ```
-@InProceedings{commonsense_QA,
-title={COMMONSENSEQA: A Question Answering Challenge Targeting Commonsense Knowledge},
-author={Alon, Talmor and Jonathan, Herzig and Nicholas, Lourie and Jonathan ,Berant},
-journal={arXiv preprint arXiv:1811.00937v2},
-year={2019}
-
+@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",
+    archivePrefix = "arXiv",
+    eprint        = "1811.00937",
+    primaryClass  = "cs",
+}
 ```
 
-
 ### Contributions
 
 Thanks to [@thomwolf](https://github.com/thomwolf), [@lewtun](https://github.com/lewtun), [@albertvillanova](https://github.com/albertvillanova), [@patrickvonplaten](https://github.com/patrickvonplaten) for adding this dataset.

commonsense_qa.py

@@ -1,4 +1,4 @@
-"""TODO(commonsense_qa): Add a description here."""
+"""CommonsenseQA dataset."""
 
 
 import json
@@ -6,111 +6,97 @@ import json
 import datasets
 
 
-# TODO(commonsense_qa): BibTeX citation
-_CITATION = """\
-@InProceedings{commonsense_QA,
-title={COMMONSENSEQA: A Question Answering Challenge Targeting Commonsense Knowledge},
-author={Alon, Talmor and Jonathan, Herzig and Nicholas, Lourie and Jonathan ,Berant},
-journal={arXiv preprint arXiv:1811.00937v2},
-year={2019}
-
-"""
+_HOMEPAGE = "https://www.tau-nlp.org/commonsenseqa"
 
-# TODO(commonsense_qa):
 _DESCRIPTION = """\
 CommonsenseQA is a new multiple-choice question answering dataset that requires different types of commonsense knowledge
- to predict the correct answers . It contains 12,102 questions with one correct answer and four distractor answers.
- The dataset is provided in two major training/validation/testing set splits: "Random split" which is the main evaluation
-  split, and "Question token split", see paper for details.
+to predict the correct answers . It contains 12,102 questions with one correct answer and four distractor answers.
+The dataset is provided in two major training/validation/testing set splits: "Random split" which is the main evaluation
+split, and "Question token split", see paper for details.
 """
 
-_URL = "https://s3.amazonaws.com/commensenseqa/"
+_CITATION = """\
+@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",
+    archivePrefix = "arXiv",
+    eprint        = "1811.00937",
+    primaryClass  = "cs",
+}
+"""
+
+_URL = "https://s3.amazonaws.com/commensenseqa"
 _URLS = {
-    "train": _URL + "train_rand_split.jsonl",
-    "dev": _URL + "dev_rand_split.jsonl",
-    "test": _URL + "test_rand_split_no_answers.jsonl",
+    "train": f"{_URL}/train_rand_split.jsonl",
+    "validation": f"{_URL}/dev_rand_split.jsonl",
+    "test": f"{_URL}/test_rand_split_no_answers.jsonl",
 }
 
 
 class CommonsenseQa(datasets.GeneratorBasedBuilder):
-    """TODO(commonsense_qa): Short description of my dataset."""
+    """CommonsenseQA dataset."""
 
-    # TODO(commonsense_qa): Set up version.
-    VERSION = datasets.Version("0.1.0")
+    VERSION = datasets.Version("1.0.0")
 
     def _info(self):
-        # These are the features of your dataset like images, labels ...
         features = datasets.Features(
             {
-                "answerKey": datasets.Value("string"),
+                "id": datasets.Value("string"),
                 "question": datasets.Value("string"),
+                "question_concept": datasets.Value("string"),
                 "choices": datasets.features.Sequence(
                     {
                         "label": datasets.Value("string"),
                         "text": datasets.Value("string"),
                     }
                 ),
+                "answerKey": datasets.Value("string"),
             }
         )
         return datasets.DatasetInfo(
-            # This is the description that will appear on the datasets page.
             description=_DESCRIPTION,
-            # datasets.features.FeatureConnectors
             features=features,
-            # If there's a common (input, target) tuple from the features,
-            # specify them here. They'll be used if as_supervised=True in
-            # builder.as_dataset.
-            supervised_keys=None,
-            # Homepage of the dataset for documentation
-            homepage="https://www.tau-datasets.org/commonsenseqa",
+            homepage=_HOMEPAGE,
             citation=_CITATION,
         )
 
     def _split_generators(self, dl_manager):
         """Returns SplitGenerators."""
-
-        download_urls = _URLS
-
-        downloaded_files = dl_manager.download_and_extract(download_urls)
-
+        filepaths = dl_manager.download_and_extract(_URLS)
+        splits = [datasets.Split.TRAIN, datasets.Split.VALIDATION, datasets.Split.TEST]
         return [
             datasets.SplitGenerator(
-                name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"], "split": "train"}
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.VALIDATION,
+                name=split,
                 gen_kwargs={
-                    "filepath": downloaded_files["dev"],
-                    "split": "dev",
+                    "filepath": filepaths[split],
                 },
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.TEST,
-                gen_kwargs={
-                    "filepath": downloaded_files["test"],
-                    "split": "test",
-                },
-            ),
+            )
+            for split in splits
         ]
 
-    def _generate_examples(self, filepath, split):
+    def _generate_examples(self, filepath):
         """Yields examples."""
-        # TODO(commonsense_qa): Yields (key, example) tuples from the dataset
         with open(filepath, encoding="utf-8") as f:
-            for id_, row in enumerate(f):
+            for uid, row in enumerate(f):
                 data = json.loads(row)
-                question = data["question"]
-                choices = question["choices"]
+                choices = data["question"]["choices"]
                 labels = [label["label"] for label in choices]
                 texts = [text["text"] for text in choices]
-                stem = question["stem"]
-                if split == "test":
-                    answerkey = ""
-                else:
-                    answerkey = data["answerKey"]
-
-                yield id_, {
-                    "answerKey": answerkey,
-                    "question": stem,
+                yield uid, {
+                    "id": data["id"],
+                    "question": data["question"]["stem"],
+                    "question_concept": data["question"]["question_concept"],
                     "choices": {"label": labels, "text": texts},
+                    "answerKey": data.get("answerKey", ""),
                 }

dataset_infos.json

@@ -1 +1 @@
-{"default": {"description": "CommonsenseQA is a new multiple-choice question answering dataset that requires different types of commonsense knowledge\n to predict the correct answers . It contains 12,102 questions with one correct answer and four distractor answers.\n The dataset is provided in two major training/validation/testing set splits: \"Random split\" which is the main evaluation\n  split, and \"Question token split\", see paper for details.\n", "citation": "@InProceedings{commonsense_QA,\ntitle={COMMONSENSEQA: A Question Answering Challenge Targeting Commonsense Knowledge},\nauthor={Alon, Talmor and Jonathan, Herzig and Nicholas, Lourie and Jonathan ,Berant},\njournal={arXiv preprint arXiv:1811.00937v2},\nyear={2019}\n\n", "homepage": "https://www.tau-nlp.org/commonsenseqa", "license": "", "features": {"answerKey": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "choices": {"feature": {"label": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "supervised_keys": null, "builder_name": "commonsense_qa", "config_name": "default", "version": {"version_str": "0.1.0", "description": null, "datasets_version_to_prepare": null, "major": 0, "minor": 1, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 205066, "num_examples": 1140, "dataset_name": "commonsense_qa"}, "train": {"name": "train", "num_bytes": 1755851, "num_examples": 9741, "dataset_name": "commonsense_qa"}, "validation": {"name": "validation", "num_bytes": 217503, "num_examples": 1221, "dataset_name": "commonsense_qa"}}, "download_checksums": {"https://s3.amazonaws.com/commensenseqa/train_rand_split.jsonl": {"num_bytes": 3785890, "checksum": "58ffa3c8472410e24b8c43f423d89c8a003d8284698a6ed7874355dedd09a2fb"}, "https://s3.amazonaws.com/commensenseqa/test_rand_split_no_answers.jsonl": {"num_bytes": 423148, "checksum": "b426896d71a9cd064cf01cfaf6e920817c51701ef66028883ac1af2e73ad5f29"}, "https://s3.amazonaws.com/commensenseqa/dev_rand_split.jsonl": {"num_bytes": 471653, "checksum": "3210497fdaae614ac085d9eb873dd7f4d49b6f965a93adadc803e1229fd8a02a"}}, "download_size": 4680691, "dataset_size": 2178420, "size_in_bytes": 6859111}}
+{"default": {"description": "CommonsenseQA is a new multiple-choice question answering dataset that requires different types of commonsense knowledge\nto predict the correct answers . It contains 12,102 questions with one correct answer and four distractor answers.\nThe dataset is provided in two major training/validation/testing set splits: \"Random split\" which is the main evaluation\nsplit, and \"Question token split\", see paper for details.\n", "citation": "@inproceedings{talmor-etal-2019-commonsenseqa,\n    title = \"{C}ommonsense{QA}: A Question Answering Challenge Targeting Commonsense Knowledge\",\n    author = \"Talmor, Alon  and\n      Herzig, Jonathan  and\n      Lourie, Nicholas  and\n      Berant, Jonathan\",\n    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)\",\n    month = jun,\n    year = \"2019\",\n    address = \"Minneapolis, Minnesota\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/N19-1421\",\n    doi = \"10.18653/v1/N19-1421\",\n    pages = \"4149--4158\",\n    archivePrefix = \"arXiv\",\n    eprint        = \"1811.00937\",\n    primaryClass  = \"cs\",\n}\n", "homepage": "https://www.tau-nlp.org/commonsenseqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "question_concept": {"dtype": "string", "id": null, "_type": "Value"}, "choices": {"feature": {"label": {"dtype": "string", "id": null, "_type": "Value"}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}, "answerKey": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "commonsense_qa", "config_name": "default", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2209044, "num_examples": 9741, "dataset_name": "commonsense_qa"}, "validation": {"name": "validation", "num_bytes": 274033, "num_examples": 1221, "dataset_name": "commonsense_qa"}, "test": {"name": "test", "num_bytes": 258017, "num_examples": 1140, "dataset_name": "commonsense_qa"}}, "download_checksums": {"https://s3.amazonaws.com/commensenseqa/train_rand_split.jsonl": {"num_bytes": 3785890, "checksum": "58ffa3c8472410e24b8c43f423d89c8a003d8284698a6ed7874355dedd09a2fb"}, "https://s3.amazonaws.com/commensenseqa/dev_rand_split.jsonl": {"num_bytes": 471653, "checksum": "3210497fdaae614ac085d9eb873dd7f4d49b6f965a93adadc803e1229fd8a02a"}, "https://s3.amazonaws.com/commensenseqa/test_rand_split_no_answers.jsonl": {"num_bytes": 423148, "checksum": "b426896d71a9cd064cf01cfaf6e920817c51701ef66028883ac1af2e73ad5f29"}}, "download_size": 4680691, "post_processing_size": null, "dataset_size": 2741094, "size_in_bytes": 7421785}}