Convert dataset to Parquet (#6)

- Convert dataset to Parquet (8c4888010de9b23a2d136ebd31d5f5ea9f3d90fa)
- Add additional data files (102463177d892baddd13db933a909a45faa02f64)
- Delete loading script (8230ac3f228b44cad2acd5038012ffed728badeb)
- Delete legacy dataset_infos.json (f7c7ec7688698d17b3876f6033a61b297fbbb0e5)

README.md

@@ -10,7 +10,6 @@ license:
 - unknown
 multilinguality:
 - monolingual
-pretty_name: OpenBookQA
 size_categories:
 - 1K<n<10K
 source_datasets:
@@ -20,8 +19,9 @@ task_categories:
 task_ids:
 - open-domain-qa
 paperswithcode_id: openbookqa
+pretty_name: OpenBookQA
 dataset_info:
-- config_name: main
+- config_name: additional
   features:
   - name: id
     dtype: string
@@ -35,19 +35,27 @@ dataset_info:
       dtype: string
   - name: answerKey
     dtype: string
+  - name: fact1
+    dtype: string
+  - name: humanScore
+    dtype: float32
+  - name: clarity
+    dtype: float32
+  - name: turkIdAnonymized
+    dtype: string
   splits:
   - name: train
-    num_bytes: 896034
+    num_bytes: 1288577
     num_examples: 4957
   - name: validation
-    num_bytes: 95519
+    num_bytes: 135916
     num_examples: 500
   - name: test
-    num_bytes: 91850
+    num_bytes: 130701
     num_examples: 500
-  download_size: 1446098
-  dataset_size: 1083403
-- config_name: additional
+  download_size: 783789
+  dataset_size: 1555194
+- config_name: main
   features:
   - name: id
     dtype: string
@@ -61,26 +69,36 @@ dataset_info:
       dtype: string
   - name: answerKey
     dtype: string
-  - name: fact1
-    dtype: string
-  - name: humanScore
-    dtype: float32
-  - name: clarity
-    dtype: float32
-  - name: turkIdAnonymized
-    dtype: string
   splits:
   - name: train
-    num_bytes: 1290473
+    num_bytes: 895386
     num_examples: 4957
   - name: validation
-    num_bytes: 136141
+    num_bytes: 95428
     num_examples: 500
   - name: test
-    num_bytes: 130926
+    num_bytes: 91759
     num_examples: 500
-  download_size: 1446098
-  dataset_size: 1557540
+  download_size: 609613
+  dataset_size: 1082573
+configs:
+- config_name: additional
+  data_files:
+  - split: train
+    path: additional/train-*
+  - split: validation
+    path: additional/validation-*
+  - split: test
+    path: additional/test-*
+- config_name: main
+  data_files:
+  - split: train
+    path: main/train-*
+  - split: validation
+    path: main/validation-*
+  - split: test
+    path: main/test-*
+  default: true
 ---
 
 # Dataset Card for OpenBookQA

additional/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33b318ea8e2354484868bc601c1b30a58149e9deb93162ff422bb8de980c7105
+size 72461

additional/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d16d719e87efb86ed0a2ac4c8cdf380f7bfb94b602088393674c0a64ce9ed3d3
+size 635446

additional/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92e5e68e4da7bec7d130d925385abf377c2d82b89a16de502b4e1b9cf3f50a26
+size 75882

dataset_infos.json

@@ -1 +0,0 @@
-{"main": {"description": "OpenBookQA aims to promote research in advanced question-answering, probing a deeper understanding of both the topic\n(with salient facts summarized as an open book, also provided with the dataset) and the language it is expressed in. In\nparticular, it contains questions that require multi-step reasoning, use of additional common and commonsense knowledge,\nand rich text comprehension.\nOpenBookQA is a new kind of question-answering dataset modeled after open book exams for assessing human understanding\nof a subject.\n", "citation": "@inproceedings{OpenBookQA2018,\n title={Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering},\n author={Todor Mihaylov and Peter Clark and Tushar Khot and Ashish Sabharwal},\n booktitle={EMNLP},\n year={2018}\n}\n", "homepage": "https://allenai.org/data/open-book-qa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question_stem": {"dtype": "string", "id": null, "_type": "Value"}, "choices": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"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": "openbookqa", "config_name": "main", "version": {"version_str": "1.0.1", "description": "", "major": 1, "minor": 0, "patch": 1}, "splits": {"train": {"name": "train", "num_bytes": 896034, "num_examples": 4957, "dataset_name": "openbookqa"}, "validation": {"name": "validation", "num_bytes": 95519, "num_examples": 500, "dataset_name": "openbookqa"}, "test": {"name": "test", "num_bytes": 91850, "num_examples": 500, "dataset_name": "openbookqa"}}, "download_checksums": {"https://s3-us-west-2.amazonaws.com/ai2-website/data/OpenBookQA-V1-Sep2018.zip": {"num_bytes": 1446098, "checksum": "82368cf05df2e3b309c17d162e10b888b4d768fad6e171e0a041954c8553be46"}}, "download_size": 1446098, "post_processing_size": null, "dataset_size": 1083403, "size_in_bytes": 2529501}, "additional": {"description": "OpenBookQA aims to promote research in advanced question-answering, probing a deeper understanding of both the topic\n(with salient facts summarized as an open book, also provided with the dataset) and the language it is expressed in. In\nparticular, it contains questions that require multi-step reasoning, use of additional common and commonsense knowledge,\nand rich text comprehension.\nOpenBookQA is a new kind of question-answering dataset modeled after open book exams for assessing human understanding\nof a subject.\n", "citation": "@inproceedings{OpenBookQA2018,\n title={Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering},\n author={Todor Mihaylov and Peter Clark and Tushar Khot and Ashish Sabharwal},\n booktitle={EMNLP},\n year={2018}\n}\n", "homepage": "https://allenai.org/data/open-book-qa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question_stem": {"dtype": "string", "id": null, "_type": "Value"}, "choices": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"dtype": "string", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}, "answerKey": {"dtype": "string", "id": null, "_type": "Value"}, "fact1": {"dtype": "string", "id": null, "_type": "Value"}, "humanScore": {"dtype": "float32", "id": null, "_type": "Value"}, "clarity": {"dtype": "float32", "id": null, "_type": "Value"}, "turkIdAnonymized": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "openbookqa", "config_name": "additional", "version": {"version_str": "1.0.1", "description": "", "major": 1, "minor": 0, "patch": 1}, "splits": {"train": {"name": "train", "num_bytes": 1290473, "num_examples": 4957, "dataset_name": "openbookqa"}, "validation": {"name": "validation", "num_bytes": 136141, "num_examples": 500, "dataset_name": "openbookqa"}, "test": {"name": "test", "num_bytes": 130926, "num_examples": 500, "dataset_name": "openbookqa"}}, "download_checksums": {"https://s3-us-west-2.amazonaws.com/ai2-website/data/OpenBookQA-V1-Sep2018.zip": {"num_bytes": 1446098, "checksum": "82368cf05df2e3b309c17d162e10b888b4d768fad6e171e0a041954c8553be46"}}, "download_size": 1446098, "post_processing_size": null, "dataset_size": 1557540, "size_in_bytes": 3003638}}

main/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd5483e366daa230c1c87bbdc512d8b7229f14f6dd04d19fc8b1a3855aaaa8a3
+size 55535

main/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98148f8a54e62eb862346a75192d5fb824d6cbb68f2f59aecd793d39ecb5cd8b
+size 495845

main/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35370b9cfee8c1ff325ccc74adc434d12c47ca0ac3244aa87f3fa77069285206
+size 58233

openbookqa.py

@@ -1,159 +0,0 @@
-"""OpenBookQA dataset."""
-
-
-import json
-import os
-import textwrap
-
-import datasets
-
-
-_HOMEPAGE = "https://allenai.org/data/open-book-qa"
-
-_DESCRIPTION = """\
-OpenBookQA aims to promote research in advanced question-answering, probing a deeper understanding of both the topic
-(with salient facts summarized as an open book, also provided with the dataset) and the language it is expressed in. In
-particular, it contains questions that require multi-step reasoning, use of additional common and commonsense knowledge,
-and rich text comprehension.
-OpenBookQA is a new kind of question-answering dataset modeled after open book exams for assessing human understanding
-of a subject.
-"""
-
-_CITATION = """\
-@inproceedings{OpenBookQA2018,
- title={Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering},
- author={Todor Mihaylov and Peter Clark and Tushar Khot and Ashish Sabharwal},
- booktitle={EMNLP},
- year={2018}
-}
-"""
-
-_URL = "https://s3-us-west-2.amazonaws.com/ai2-website/data/OpenBookQA-V1-Sep2018.zip"
-
-
-class OpenbookqaConfig(datasets.BuilderConfig):
-    def __init__(self, data_dir=None, filenames=None, version=datasets.Version("1.0.1", ""), **kwargs):
-        """BuilderConfig for openBookQA dataset
-
-        Args:
-          data_dir: directory for the given dataset name
-          **kwargs: keyword arguments forwarded to super.
-        """
-        super().__init__(version=version, **kwargs)
-        self.data_dir = data_dir
-        self.filenames = filenames
-
-
-class Openbookqa(datasets.GeneratorBasedBuilder):
-    """OpenBookQA dataset."""
-
-    BUILDER_CONFIGS = [
-        OpenbookqaConfig(
-            name="main",
-            description=textwrap.dedent(
-                """\
-                It consists of 5,957 multiple-choice elementary-level science questions (4,957 train, 500 dev, 500 test),
-                which probe the understanding of a small “book” of 1,326 core science facts and the application of these facts to novel
-                situations. For training, the dataset includes a mapping from each question to the core science fact it was designed to
-                probe. Answering OpenBookQA questions requires additional broad common knowledge, not contained in the book. The questions,
-                by design, are answered incorrectly by both a retrieval-based algorithm and a word co-occurrence algorithm. Strong neural
-                baselines achieve around 50% on OpenBookQA, leaving a large gap to the 92% accuracy of crowd-workers.
-                """
-            ),
-            data_dir="Main",
-            filenames={
-                "train": "train.jsonl",
-                "validation": "dev.jsonl",
-                "test": "test.jsonl",
-            },
-        ),
-        OpenbookqaConfig(
-            name="additional",
-            description=textwrap.dedent(
-                """\
-                Additionally, we provide 5,167 crowd-sourced common knowledge facts, and an expanded version of the train/dev/test questions where
-                each question is associated with its originating core fact, a human accuracy score, a clarity score, and an anonymized crowd-worker
-                ID (in the 'Additional' folder).
-                """
-            ),
-            data_dir="Additional",
-            filenames={
-                "train": "train_complete.jsonl",
-                "validation": "dev_complete.jsonl",
-                "test": "test_complete.jsonl",
-            },
-        ),
-    ]
-    DEFAULT_CONFIG_NAME = "main"
-
-    def _info(self):
-        if self.config.name == "main":
-            features = datasets.Features(
-                {
-                    "id": datasets.Value("string"),
-                    "question_stem": datasets.Value("string"),
-                    "choices": datasets.features.Sequence(
-                        {
-                            "text": datasets.Value("string"),
-                            "label": datasets.Value("string"),
-                        }
-                    ),
-                    "answerKey": datasets.Value("string"),
-                }
-            )
-        else:
-            features = datasets.Features(
-                {
-                    "id": datasets.Value("string"),
-                    "question_stem": datasets.Value("string"),
-                    "choices": datasets.features.Sequence(
-                        {
-                            "text": datasets.Value("string"),
-                            "label": datasets.Value("string"),
-                        }
-                    ),
-                    "answerKey": datasets.Value("string"),
-                    "fact1": datasets.Value("string"),
-                    "humanScore": datasets.Value("float"),
-                    "clarity": datasets.Value("float"),
-                    "turkIdAnonymized": datasets.Value("string"),
-                }
-            )
-        return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=features,
-            homepage=_HOMEPAGE,
-            citation=_CITATION,
-        )
-
-    def _split_generators(self, dl_manager):
-        """Returns SplitGenerators."""
-        dl_dir = dl_manager.download_and_extract(_URL)
-        data_dir = os.path.join(dl_dir, "OpenBookQA-V1-Sep2018", "Data", self.config.data_dir)
-        splits = [datasets.Split.TRAIN, datasets.Split.VALIDATION, datasets.Split.TEST]
-        return [
-            datasets.SplitGenerator(
-                name=split,
-                gen_kwargs={"filepath": os.path.join(data_dir, self.config.filenames[split])},
-            )
-            for split in splits
-        ]
-
-    def _generate_examples(self, filepath):
-        """Yields examples."""
-        with open(filepath, encoding="utf-8") as f:
-            for uid, row in enumerate(f):
-                data = json.loads(row)
-                example = {
-                    "id": data["id"],
-                    "question_stem": data["question"]["stem"],
-                    "choices": {
-                        "text": [choice["text"] for choice in data["question"]["choices"]],
-                        "label": [choice["label"] for choice in data["question"]["choices"]],
-                    },
-                    "answerKey": data["answerKey"],
-                }
-                if self.config.name == "additional":
-                    for key in ["fact1", "humanScore", "clarity", "turkIdAnonymized"]:
-                        example[key] = data[key]
-                yield uid, example