Convert dataset to Parquet (#2)

- Convert dataset to Parquet (76b601c821b066397098c65c1ac75b10c6f11949)
- Delete loading script (1da014ce7ff4fd6edb88d0fef62bbac63e867259)
- Delete legacy dataset_infos.json (319e71e380040f70c98a21661a5bc174ecd80c6b)

README.md

@@ -67,17 +67,26 @@ dataset_info:
     - name: less_cause_dir
       dtype: string
   splits:
-  - name: test
-    num_bytes: 351374
-    num_examples: 784
   - name: train
-    num_bytes: 1197525
+    num_bytes: 1188342
     num_examples: 2696
+  - name: test
+    num_bytes: 348644
+    num_examples: 784
   - name: validation
-    num_bytes: 175871
+    num_bytes: 174491
     num_examples: 384
-  download_size: 497354
-  dataset_size: 1724770
+  download_size: 569255
+  dataset_size: 1711477
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: test
+    path: data/test-*
+  - split: validation
+    path: data/validation-*
 ---
 
 # Dataset Card for "quartz"

data/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78c4c77cfc1ad0e65dbdfeec4ed788117e73484ad44f6a2aebf43e08c7fba2e8
+size 99764

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edeb1fa716d945ee198c9f5ce5ab6cad4d847e714ff7463834dee422c8450035
+size 415195

data/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91afc8bbdb1b6b67f35f05ea4b4dbc1ba718ec2469be665a386c46b4648a1fb4
+size 54296

dataset_infos.json

@@ -1 +0,0 @@
-{"default": {"description": "QuaRTz is a crowdsourced dataset of 3864 multiple-choice questions about open domain qualitative relationships. Each \nquestion is paired with one of 405 different background sentences (sometimes short paragraphs).\nThe QuaRTz dataset V1 contains 3864 questions about open domain qualitative relationships. Each question is paired with \none of 405 different background sentences (sometimes short paragraphs).\n\nThe dataset is split into train (2696), dev (384) and test (784). A background sentence will only appear in a single split.\n", "citation": "@InProceedings{quartz,\n  author = {Oyvind Tafjord and Matt Gardner and Kevin Lin and Peter Clark},\n  title = {\"QUARTZ: An Open-Domain Dataset of Qualitative Relationship\nQuestions\"},\n \n  year = {\"2019\"},\n}\n", "homepage": "https://allenai.org/data/quartz", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"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"}, "para": {"dtype": "string", "id": null, "_type": "Value"}, "para_id": {"dtype": "string", "id": null, "_type": "Value"}, "para_anno": {"effect_prop": {"dtype": "string", "id": null, "_type": "Value"}, "cause_dir_str": {"dtype": "string", "id": null, "_type": "Value"}, "effect_dir_str": {"dtype": "string", "id": null, "_type": "Value"}, "cause_dir_sign": {"dtype": "string", "id": null, "_type": "Value"}, "effect_dir_sign": {"dtype": "string", "id": null, "_type": "Value"}, "cause_prop": {"dtype": "string", "id": null, "_type": "Value"}}, "question_anno": {"more_effect_dir": {"dtype": "string", "id": null, "_type": "Value"}, "less_effect_dir": {"dtype": "string", "id": null, "_type": "Value"}, "less_cause_prop": {"dtype": "string", "id": null, "_type": "Value"}, "more_effect_prop": {"dtype": "string", "id": null, "_type": "Value"}, "less_effect_prop": {"dtype": "string", "id": null, "_type": "Value"}, "less_cause_dir": {"dtype": "string", "id": null, "_type": "Value"}}}, "supervised_keys": null, "builder_name": "quartz", "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": 351374, "num_examples": 784, "dataset_name": "quartz"}, "train": {"name": "train", "num_bytes": 1197525, "num_examples": 2696, "dataset_name": "quartz"}, "validation": {"name": "validation", "num_bytes": 175871, "num_examples": 384, "dataset_name": "quartz"}}, "download_checksums": {"https://s3-us-west-2.amazonaws.com/ai2-website/data/quartz-dataset-v1-aug2019.zip": {"num_bytes": 497354, "checksum": "e86ed35153c6c3fb6dc5991b6a3b520a2c154c42266cb6b4edc7ed526fa4b5a8"}}, "download_size": 497354, "dataset_size": 1724770, "size_in_bytes": 2222124}}

quartz.py

@@ -1,158 +0,0 @@
-"""TODO(quartz): Add a description here."""
-
-
-import json
-import os
-
-import datasets
-
-
-# TODO(quartz): BibTeX citation
-_CITATION = """\
-@InProceedings{quartz,
-  author = {Oyvind Tafjord and Matt Gardner and Kevin Lin and Peter Clark},
-  title = {"QUARTZ: An Open-Domain Dataset of Qualitative Relationship
-Questions"},
-  year = {"2019"},
-}
-"""
-
-# TODO(quartz):
-_DESCRIPTION = """\
-QuaRTz is a crowdsourced dataset of 3864 multiple-choice questions about open domain qualitative relationships. Each
-question is paired with one of 405 different background sentences (sometimes short paragraphs).
-The QuaRTz dataset V1 contains 3864 questions about open domain qualitative relationships. Each question is paired with
-one of 405 different background sentences (sometimes short paragraphs).
-The dataset is split into train (2696), dev (384) and test (784). A background sentence will only appear in a single split.
-"""
-
-_URL = "https://s3-us-west-2.amazonaws.com/ai2-website/data/quartz-dataset-v1-aug2019.zip"
-
-
-class Quartz(datasets.GeneratorBasedBuilder):
-    """TODO(quartz): Short description of my dataset."""
-
-    # TODO(quartz): Set up version.
-    VERSION = datasets.Version("0.1.0")
-
-    def _info(self):
-        # TODO(quartz): Specifies the datasets.DatasetInfo object
-        return datasets.DatasetInfo(
-            # This is the description that will appear on the datasets page.
-            description=_DESCRIPTION,
-            # datasets.features.FeatureConnectors
-            features=datasets.Features(
-                {
-                    # These are the features of your dataset like images, labels ...
-                    "id": datasets.Value("string"),
-                    "question": datasets.Value("string"),
-                    "choices": datasets.features.Sequence(
-                        {"text": datasets.Value("string"), "label": datasets.Value("string")}
-                    ),
-                    "answerKey": datasets.Value("string"),
-                    "para": datasets.Value("string"),
-                    "para_id": datasets.Value("string"),
-                    "para_anno": {
-                        "effect_prop": datasets.Value("string"),
-                        "cause_dir_str": datasets.Value("string"),
-                        "effect_dir_str": datasets.Value("string"),
-                        "cause_dir_sign": datasets.Value("string"),
-                        "effect_dir_sign": datasets.Value("string"),
-                        "cause_prop": datasets.Value("string"),
-                    },
-                    "question_anno": {
-                        "more_effect_dir": datasets.Value("string"),
-                        "less_effect_dir": datasets.Value("string"),
-                        "less_cause_prop": datasets.Value("string"),
-                        "more_effect_prop": datasets.Value("string"),
-                        "less_effect_prop": datasets.Value("string"),
-                        "less_cause_dir": datasets.Value("string"),
-                    },
-                }
-            ),
-            # 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://allenai.org/data/quartz",
-            citation=_CITATION,
-        )
-
-    def _split_generators(self, dl_manager):
-        """Returns SplitGenerators."""
-        # TODO(quartz): Downloads the data and defines the splits
-        # dl_manager is a datasets.download.DownloadManager that can be used to
-        # download and extract URLs
-        dl_dir = dl_manager.download_and_extract(_URL)
-        data_dir = os.path.join(dl_dir, "quartz-dataset-v1-aug2019")
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                # These kwargs will be passed to _generate_examples
-                gen_kwargs={"filepath": os.path.join(data_dir, "train.jsonl")},
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.TEST,
-                # These kwargs will be passed to _generate_examples
-                gen_kwargs={"filepath": os.path.join(data_dir, "test.jsonl")},
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.VALIDATION,
-                # These kwargs will be passed to _generate_examples
-                gen_kwargs={"filepath": os.path.join(data_dir, "dev.jsonl")},
-            ),
-        ]
-
-    def _generate_examples(self, filepath):
-        """Yields examples."""
-        # TODO(quartz): Yields (key, example) tuples from the dataset
-        with open(filepath, encoding="utf-8") as f:
-            for row in f:
-                data = json.loads(row)
-                id_ = data["id"]
-                question = data["question"]["stem"]
-                answerKey = data["answerKey"]
-                choices = data["question"]["choices"]
-                choice_text = [choice["text"] for choice in choices]
-                choice_label = [choice["label"] for choice in choices]
-                para_id = data["para_id"]
-                para = data["para"]
-                para_ano = data["para_anno"]
-                effect_prop = para_ano.get("effect_prop", "")
-                cause_dir_str = para_ano.get("cause_dir_str", "")
-                effect_dir_str = para_ano.get("effect_dir_str", "")
-                cause_dir_sign = para_ano.get("cause_dir_sign", "")
-                effect_dir_sign = para_ano.get("effect_dir_sign", "")
-                cause_prop = para_ano.get("cause_prop", "")
-                question_anno = data["question_anno"]
-                more_effect_dir = "" if not question_anno else question_anno.get("more_effect_dir", "")
-                less_effect_dir = "" if not question_anno else question_anno.get("less_effect_dir", "")
-                less_cause_prop = "" if not question_anno else question_anno.get("less_cause_prop", "")
-                more_effect_prop = "" if not question_anno else question_anno.get("more_effect_prop", "")
-                less_effect_prop = "" if not question_anno else question_anno.get("less_effect_prop", "")
-                less_cause_dir = "" if not question_anno else question_anno.get("less_effect_prop", "")
-                yield id_, {
-                    "id": id_,
-                    "question": question,
-                    "choices": {"text": choice_text, "label": choice_label},
-                    "answerKey": answerKey,
-                    "para": para,
-                    "para_id": para_id,
-                    "para_anno": {
-                        "effect_prop": effect_prop,
-                        "cause_dir_str": cause_dir_str,
-                        "effect_dir_str": effect_dir_str,
-                        "cause_dir_sign": cause_dir_sign,
-                        "effect_dir_sign": effect_dir_sign,
-                        "cause_prop": cause_prop,
-                    },
-                    "question_anno": {
-                        "more_effect_dir": more_effect_dir,
-                        "less_effect_dir": less_effect_dir,
-                        "less_cause_prop": less_cause_prop,
-                        "more_effect_prop": more_effect_prop,
-                        "less_effect_prop": less_effect_prop,
-                        "less_cause_dir": less_cause_dir,
-                    },
-                }