Convert dataset to Parquet

README.md

@@ -42,8 +42,15 @@ dataset_info:
   - name: validation
     num_bytes: 168004403
     num_examples: 930062
-  download_size: 1482064429
+  download_size: 2496835395
   dataset_size: 3824869475
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: validation
+    path: data/validation-*
 ---
 
 # Dataset Card for "asnq"

asnq.py

@@ -1,151 +0,0 @@
-# coding=utf-8
-# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-"""Answer-Sentence Natural Questions (ASNQ)
-
-ASNQ is a dataset for answer sentence selection derived from Google's
-Natural Questions (NQ) dataset (Kwiatkowski et al. 2019). It converts
-NQ's dataset into an AS2 (answer-sentence-selection) format.
-
-The dataset details can be found in the paper at
-https://arxiv.org/abs/1911.04118
-
-The dataset can be downloaded at
-https://d3t7erp6ge410c.cloudfront.net/tanda-aaai-2020/data/asnq.tar
-
-"""
-
-
-import csv
-import os
-
-import datasets
-
-
-_CITATION = """\
-@article{garg2019tanda,
-    title={TANDA: Transfer and Adapt Pre-Trained Transformer Models for Answer Sentence Selection},
-    author={Siddhant Garg and Thuy Vu and Alessandro Moschitti},
-    year={2019},
-    eprint={1911.04118},
-}
-"""
-
-_DESCRIPTION = """\
-ASNQ is a dataset for answer sentence selection derived from
-Google's Natural Questions (NQ) dataset (Kwiatkowski et al. 2019).
-
-Each example contains a question, candidate sentence, label indicating whether or not
-the sentence answers the question, and two additional features --
-sentence_in_long_answer and short_answer_in_sentence indicating whether ot not the
-candidate sentence is contained in the long_answer and if the short_answer is in the candidate sentence.
-
-For more details please see
-https://arxiv.org/pdf/1911.04118.pdf
-
-and
-
-https://research.google/pubs/pub47761/
-"""
-
-_URL = "data/asnq.zip"
-
-
-class ASNQ(datasets.GeneratorBasedBuilder):
-    """ASNQ is a dataset for answer sentence selection derived
-    ASNQ is a dataset for answer sentence selection derived from
-    Google's Natural Questions (NQ) dataset (Kwiatkowski et al. 2019).
-
-    The dataset details can be found in the paper:
-    https://arxiv.org/abs/1911.04118
-    """
-
-    VERSION = datasets.Version("1.0.0")
-
-    def _info(self):
-
-        return datasets.DatasetInfo(
-            # This is the description that will appear on the datasets page.
-            description=_DESCRIPTION,
-            # This defines the different columns of the dataset and their types
-            features=datasets.Features(
-                {
-                    "question": datasets.Value("string"),
-                    "sentence": datasets.Value("string"),
-                    "label": datasets.ClassLabel(names=["neg", "pos"]),
-                    "sentence_in_long_answer": datasets.Value("bool"),
-                    "short_answer_in_sentence": datasets.Value("bool"),
-                }
-            ),
-            # No default supervised_keys
-            supervised_keys=None,
-            # Homepage of the dataset for documentation
-            homepage="https://github.com/alexa/wqa_tanda#answer-sentence-natural-questions-asnq",
-            citation=_CITATION,
-        )
-
-    def _split_generators(self, dl_manager):
-        """Returns SplitGenerators."""
-        # 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, "data", "asnq")
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                # These kwargs will be passed to _generate_examples
-                gen_kwargs={
-                    "filepath": os.path.join(data_dir, "train.tsv"),
-                    "split": "train",
-                },
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.VALIDATION,
-                # These kwargs will be passed to _generate_examples
-                gen_kwargs={
-                    "filepath": os.path.join(data_dir, "dev.tsv"),
-                    "split": "dev",
-                },
-            ),
-        ]
-
-    def _generate_examples(self, filepath, split):
-        """Yields examples.
-
-        Original dataset contains labels '1', '2', '3' and '4', with labels
-        '1', '2' and '3' considered negative (sentence does not answer the question),
-        and label '4' considered positive (sentence does answer the question).
-        We map these labels to two classes, returning the other properties as additional
-        features."""
-
-        # Mapping of dataset's original labels to a tuple of
-        # (label, sentence_in_long_answer, short_answer_in_sentence)
-        label_map = {
-            "1": ("neg", False, False),
-            "2": ("neg", False, True),
-            "3": ("neg", True, False),
-            "4": ("pos", True, True),
-        }
-        with open(filepath, encoding="utf-8") as tsvfile:
-            tsvreader = csv.reader(tsvfile, delimiter="\t")
-            for id_, row in enumerate(tsvreader):
-                question, sentence, orig_label = row
-                label, sentence_in_long_answer, short_answer_in_sentence = label_map[orig_label]
-                yield id_, {
-                    "question": question,
-                    "sentence": sentence,
-                    "label": label,
-                    "sentence_in_long_answer": sentence_in_long_answer,
-                    "short_answer_in_sentence": short_answer_in_sentence,
-                }

data/{asnq.zip → train-00000-of-00008.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4d1b3ee889be3d2a090d075a7494cee339f6fcfae6bc16ce5d06ad66874e591c
-size 1482064429
+oid sha256:c51e6cd0278c630e2094d9cdd5502fb3f35e60e260477a969832b8422cf5574e
+size 299139770

data/train-00001-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b74a1689c47410893d8e944a994bd8e610ab173cb9e9255e225d1777b7d0859a
+size 298212695

data/train-00002-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9dd2783190464db93960c13fd8e99e949b52add2bdb4b5a9267e8601cfaef53
+size 299688447

data/train-00003-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7ff76549b7fc5e055419e274a1b28a277ac2742ed15298f47ee287530323407
+size 299173033

data/train-00004-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1baa386b5d338dea200eec4aebf3f553bf5591b059a0efbf587857c8713f3e95
+size 299751866

data/train-00005-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15c4c802114cfd6db510dbcecd5155c58b92758b06c01fd91fe94bfcb411b9cb
+size 299035267

data/train-00006-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce69b66214e21c3330a1836e652742876615e462e59b973fdfd89ee21c38303b
+size 299086683

data/train-00007-of-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b090f7299765d9101cc87ec2b29b05b13e96acd6e6717c91282b06f0e606e567
+size 299236601

data/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c400a3e7bbd1b1f184bbc3bda3747b3946ccfa29e76c7ce2841613601515fed6
+size 103511033

dataset_infos.json

@@ -1 +0,0 @@
-{"default": {"description": "ASNQ is a dataset for answer sentence selection derived from\nGoogle's Natural Questions (NQ) dataset (Kwiatkowski et al. 2019).\n\nEach example contains a question, candidate sentence, label indicating whether or not\nthe sentence answers the question, and two additional features --\nsentence_in_long_answer and short_answer_in_sentence indicating whether ot not the\ncandidate sentence is contained in the long_answer and if the short_answer is in the candidate sentence.\n\nFor more details please see\nhttps://arxiv.org/pdf/1911.04118.pdf\n\nand\n\nhttps://research.google/pubs/pub47761/\n", "citation": "@article{garg2019tanda,\n    title={TANDA: Transfer and Adapt Pre-Trained Transformer Models for Answer Sentence Selection},\n    author={Siddhant Garg and Thuy Vu and Alessandro Moschitti},\n    year={2019},\n    eprint={1911.04118},\n}\n", "homepage": "https://github.com/alexa/wqa_tanda#answer-sentence-natural-questions-asnq", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "sentence": {"dtype": "string", "id": null, "_type": "Value"}, "label": {"num_classes": 2, "names": ["neg", "pos"], "id": null, "_type": "ClassLabel"}, "sentence_in_long_answer": {"dtype": "bool", "id": null, "_type": "Value"}, "short_answer_in_sentence": {"dtype": "bool", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "asnq", "config_name": "default", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3656865072, "num_examples": 20377568, "dataset_name": "asnq"}, "validation": {"name": "validation", "num_bytes": 168004403, "num_examples": 930062, "dataset_name": "asnq"}}, "download_checksums": {"data/asnq.zip": {"num_bytes": 1482064429, "checksum": "4d1b3ee889be3d2a090d075a7494cee339f6fcfae6bc16ce5d06ad66874e591c"}}, "download_size": 1482064429, "post_processing_size": null, "dataset_size": 3824869475, "size_in_bytes": 5306933904}}