Add wikitablequestions dataset (#3870)

* Add wikitablequestions dataset

* Using tsv instead of csv file to support better.

* fix checksum for dataset wikitablequestions - pass all tests.

* Fix the answer as a sequence instead of a string.

* reduce the dummy data size

* fix the answer name and the table example json

* Fix the answer as a sequence instead of a string.

* Fix the dummy data files.

* * fix the skip on streaming mode.

* * remove other dummy data

Commit from https://github.com/huggingface/datasets/commit/b2af98ca83f4509a0c885c3187bfe97f38c9d99c

README.md

@@ -0,0 +1,189 @@
+---
+annotations_creators:
+- crowdsourced
+language_creators:
+- found
+languages:
+- en
+licenses:
+- cc-by-4-0
+multilinguality:
+- monolingual
+paperswithcode_id: null
+pretty_name: WikiTableQuestions
+size_categories:
+- 10K<n<100K
+source_datasets:
+- original
+task_categories:
+- question-answering
+task_ids:
+- question-answering-other-table-question-answering
+---
+
+# Dataset Card for WikiTableQuestions
+
+## Table of Contents
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-instances)
+  - [Data Splits](#data-instances)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+
+## Dataset Description
+
+- **Homepage:** [WikiTableQuestions homepage](https://nlp.stanford.edu/software/sempre/wikitable)
+- **Repository:** [WikiTableQuestions repository](https://github.com/ppasupat/WikiTableQuestions)
+- **Paper:** [Compositional Semantic Parsing on Semi-Structured Tables](https://arxiv.org/abs/1508.00305)
+- **Leaderboard:** [WikiTableQuestions leaderboard on PaperWithCode](https://paperswithcode.com/dataset/wikitablequestions)
+- **Point of Contact:** [Needs More Information]
+
+### Dataset Summary
+
+The WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.
+
+### Supported Tasks and Leaderboards
+
+question-answering, table-question-answering
+
+### Languages
+
+en
+
+## Dataset Structure
+
+### Data Instances
+
+#### default
+
+- **Size of downloaded dataset files:** 27.91 MB
+- **Size of the generated dataset:** 45.68 MB
+- **Total amount of disk used:** 73.60 MB
+
+An example of 'validation' looks as follows:
+```
+{
+    "id": "nt-0",
+    "question": "what was the last year where this team was a part of the usl a-league?",
+    "answers": ["2004"],
+    "table": {
+        "header": ["Year", "Division", "League", ...], 
+        "name": "csv/204-csv/590.csv", 
+        "rows": [
+           ["2001", "2", "USL A-League", ...],
+           ["2002", "2", "USL A-League", ...], 
+           ...
+        ]
+    }
+}
+```
+
+### Data Fields
+
+The data fields are the same among all splits.
+
+#### default
+- `id`: a `string` feature.
+- `question`: a `string` feature.
+- `answers`: a `list` of `string` feature.
+- `table`: a dictionary feature containing:
+  - `header`: a `list` of `string` features.
+  - `rows`: a `list` of `list` of `string` features:
+  - `name`: a `string` feature.
+
+### Data Splits
+
+| name  |train|validation|test |
+|-------|----:|---------:|----:|
+|default|11321|      2831|4344|
+
+## Dataset Creation
+
+### Curation Rationale
+
+[Needs More Information]
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+[Needs More Information]
+
+#### Who are the source language producers?
+
+[Needs More Information]
+
+### Annotations
+
+#### Annotation process
+
+[Needs More Information]
+
+#### Who are the annotators?
+
+[Needs More Information]
+
+### Personal and Sensitive Information
+
+[Needs More Information]
+
+## Considerations for Using the Data
+
+### Social Impact of Dataset
+
+[Needs More Information]
+
+### Discussion of Biases
+
+[Needs More Information]
+
+### Other Known Limitations
+
+[Needs More Information]
+
+## Additional Information
+
+### Dataset Curators
+
+Panupong Pasupat and Percy Liang
+
+### Licensing Information
+
+Creative Commons Attribution Share Alike 4.0 International
+
+### Citation Information
+
+```
+@inproceedings{pasupat-liang-2015-compositional,
+    title = "Compositional Semantic Parsing on Semi-Structured Tables",
+    author = "Pasupat, Panupong and Liang, Percy",
+    booktitle = "Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
+    month = jul,
+    year = "2015",
+    address = "Beijing, China",
+    publisher = "Association for Computational Linguistics",
+    url = "https://aclanthology.org/P15-1142",
+    doi = "10.3115/v1/P15-1142",
+    pages = "1470--1480",
+}
+```
+
+### Contributions
+
+Thanks to [@SivilTaram](https://github.com/SivilTaram) for adding this dataset.

dataset_infos.json

@@ -0,0 +1 @@
+{"random-split-1": {"description": "This WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.\n", "citation": "@inproceedings{pasupat-liang-2015-compositional,\n    title = \"Compositional Semantic Parsing on Semi-Structured Tables\",\n    author = \"Pasupat, Panupong and Liang, Percy\",\n    booktitle = \"Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)\",\n    month = jul,\n    year = \"2015\",\n    address = \"Beijing, China\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/P15-1142\",\n    doi = \"10.3115/v1/P15-1142\",\n    pages = \"1470--1480\",\n}\n", "homepage": "https://nlp.stanford.edu/software/sempre/wikitable", "license": "Creative Commons Attribution Share Alike 4.0 International", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "table": {"header": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "rows": {"feature": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "length": -1, "id": null, "_type": "Sequence"}, "name": {"dtype": "string", "id": null, "_type": "Value"}}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "wiki_table_questions", "config_name": "random-split-1", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 30364389, "num_examples": 11321, "dataset_name": "wiki_table_questions"}, "test": {"name": "test", "num_bytes": 11423506, "num_examples": 4344, "dataset_name": "wiki_table_questions"}, "validation": {"name": "validation", "num_bytes": 7145768, "num_examples": 2831, "dataset_name": "wiki_table_questions"}}, "download_checksums": {"https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip": {"num_bytes": 29267445, "checksum": "7c9ca7cc1ccd75fe4be0255b44be63f7b566761005f4ee6ce67e51c129d8b085"}}, "download_size": 29267445, "post_processing_size": null, "dataset_size": 48933663, "size_in_bytes": 78201108}, "random-split-2": {"description": "This WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.\n", "citation": "@inproceedings{pasupat-liang-2015-compositional,\n    title = \"Compositional Semantic Parsing on Semi-Structured Tables\",\n    author = \"Pasupat, Panupong and Liang, Percy\",\n    booktitle = \"Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)\",\n    month = jul,\n    year = \"2015\",\n    address = \"Beijing, China\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/P15-1142\",\n    doi = \"10.3115/v1/P15-1142\",\n    pages = \"1470--1480\",\n}\n", "homepage": "https://nlp.stanford.edu/software/sempre/wikitable", "license": "Creative Commons Attribution Share Alike 4.0 International", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "table": {"header": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "rows": {"feature": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "length": -1, "id": null, "_type": "Sequence"}, "name": {"dtype": "string", "id": null, "_type": "Value"}}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "wiki_table_questions", "config_name": "random-split-2", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 30098954, "num_examples": 11314, "dataset_name": "wiki_table_questions"}, "test": {"name": "test", "num_bytes": 11423506, "num_examples": 4344, "dataset_name": "wiki_table_questions"}, "validation": {"name": "validation", "num_bytes": 7411203, "num_examples": 2838, "dataset_name": "wiki_table_questions"}}, "download_checksums": {"https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip": {"num_bytes": 29267445, "checksum": "7c9ca7cc1ccd75fe4be0255b44be63f7b566761005f4ee6ce67e51c129d8b085"}}, "download_size": 29267445, "post_processing_size": null, "dataset_size": 48933663, "size_in_bytes": 78201108}, "random-split-3": {"description": "This WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.\n", "citation": "@inproceedings{pasupat-liang-2015-compositional,\n    title = \"Compositional Semantic Parsing on Semi-Structured Tables\",\n    author = \"Pasupat, Panupong and Liang, Percy\",\n    booktitle = \"Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)\",\n    month = jul,\n    year = \"2015\",\n    address = \"Beijing, China\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/P15-1142\",\n    doi = \"10.3115/v1/P15-1142\",\n    pages = \"1470--1480\",\n}\n", "homepage": "https://nlp.stanford.edu/software/sempre/wikitable", "license": "Creative Commons Attribution Share Alike 4.0 International", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "table": {"header": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "rows": {"feature": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "length": -1, "id": null, "_type": "Sequence"}, "name": {"dtype": "string", "id": null, "_type": "Value"}}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "wiki_table_questions", "config_name": "random-split-3", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 28778697, "num_examples": 11314, "dataset_name": "wiki_table_questions"}, "test": {"name": "test", "num_bytes": 11423506, "num_examples": 4344, "dataset_name": "wiki_table_questions"}, "validation": {"name": "validation", "num_bytes": 8731460, "num_examples": 2838, "dataset_name": "wiki_table_questions"}}, "download_checksums": {"https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip": {"num_bytes": 29267445, "checksum": "7c9ca7cc1ccd75fe4be0255b44be63f7b566761005f4ee6ce67e51c129d8b085"}}, "download_size": 29267445, "post_processing_size": null, "dataset_size": 48933663, "size_in_bytes": 78201108}, "random-split-4": {"description": "This WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.\n", "citation": "@inproceedings{pasupat-liang-2015-compositional,\n    title = \"Compositional Semantic Parsing on Semi-Structured Tables\",\n    author = \"Pasupat, Panupong and Liang, Percy\",\n    booktitle = \"Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)\",\n    month = jul,\n    year = \"2015\",\n    address = \"Beijing, China\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/P15-1142\",\n    doi = \"10.3115/v1/P15-1142\",\n    pages = \"1470--1480\",\n}\n", "homepage": "https://nlp.stanford.edu/software/sempre/wikitable", "license": "Creative Commons Attribution Share Alike 4.0 International", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "table": {"header": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "rows": {"feature": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "length": -1, "id": null, "_type": "Sequence"}, "name": {"dtype": "string", "id": null, "_type": "Value"}}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "wiki_table_questions", "config_name": "random-split-4", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 30166421, "num_examples": 11321, "dataset_name": "wiki_table_questions"}, "test": {"name": "test", "num_bytes": 11423506, "num_examples": 4344, "dataset_name": "wiki_table_questions"}, "validation": {"name": "validation", "num_bytes": 7343736, "num_examples": 2831, "dataset_name": "wiki_table_questions"}}, "download_checksums": {"https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip": {"num_bytes": 29267445, "checksum": "7c9ca7cc1ccd75fe4be0255b44be63f7b566761005f4ee6ce67e51c129d8b085"}}, "download_size": 29267445, "post_processing_size": null, "dataset_size": 48933663, "size_in_bytes": 78201108}, "random-split-5": {"description": "This WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.\n", "citation": "@inproceedings{pasupat-liang-2015-compositional,\n    title = \"Compositional Semantic Parsing on Semi-Structured Tables\",\n    author = \"Pasupat, Panupong and Liang, Percy\",\n    booktitle = \"Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)\",\n    month = jul,\n    year = \"2015\",\n    address = \"Beijing, China\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://aclanthology.org/P15-1142\",\n    doi = \"10.3115/v1/P15-1142\",\n    pages = \"1470--1480\",\n}\n", "homepage": "https://nlp.stanford.edu/software/sempre/wikitable", "license": "Creative Commons Attribution Share Alike 4.0 International", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "table": {"header": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "rows": {"feature": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "length": -1, "id": null, "_type": "Sequence"}, "name": {"dtype": "string", "id": null, "_type": "Value"}}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "wiki_table_questions", "config_name": "random-split-5", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 30333964, "num_examples": 11316, "dataset_name": "wiki_table_questions"}, "test": {"name": "test", "num_bytes": 11423506, "num_examples": 4344, "dataset_name": "wiki_table_questions"}, "validation": {"name": "validation", "num_bytes": 7176193, "num_examples": 2836, "dataset_name": "wiki_table_questions"}}, "download_checksums": {"https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip": {"num_bytes": 29267445, "checksum": "7c9ca7cc1ccd75fe4be0255b44be63f7b566761005f4ee6ce67e51c129d8b085"}}, "download_size": 29267445, "post_processing_size": null, "dataset_size": 48933663, "size_in_bytes": 78201108}}

dummy/random-split-1/1.0.2/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:117d768167b66026421067e7ca8c9ad2bc5d32a4c7e622e77d5fbbb2843c4ef5
+size 56682

wikitablequestions.py

@@ -0,0 +1,184 @@
+# 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.
+"""The WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables."""
+
+import os
+
+import datasets
+
+
+# Find for instance the citation on arxiv or on the dataset repo/website
+_CITATION = """\
+@inproceedings{pasupat-liang-2015-compositional,
+    title = "Compositional Semantic Parsing on Semi-Structured Tables",
+    author = "Pasupat, Panupong and Liang, Percy",
+    booktitle = "Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
+    month = jul,
+    year = "2015",
+    address = "Beijing, China",
+    publisher = "Association for Computational Linguistics",
+    url = "https://aclanthology.org/P15-1142",
+    doi = "10.3115/v1/P15-1142",
+    pages = "1470--1480",
+}
+"""
+
+# You can copy an official description
+_DESCRIPTION = """\
+This WikiTableQuestions dataset is a large-scale dataset for the task of question answering on semi-structured tables.
+"""
+
+_HOMEPAGE = "https://nlp.stanford.edu/software/sempre/wikitable"
+
+_LICENSE = "Creative Commons Attribution Share Alike 4.0 International"
+
+# The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
+# This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
+_DATA_URL = (
+    "https://github.com/ppasupat/WikiTableQuestions/releases/download/v1.0.2/WikiTableQuestions-1.0.2-compact.zip"
+)
+
+
+class WikiTableQuestions(datasets.GeneratorBasedBuilder):
+    """WikiTableQuestions: a large-scale dataset for the task of question answering on semi-structured tables."""
+
+    VERSION = datasets.Version("1.0.2")
+
+    # This is an example of a dataset with multiple configurations.
+    # If you don't want/need to define several sub-sets in your dataset,
+    # just remove the BUILDER_CONFIG_CLASS and the BUILDER_CONFIGS attributes.
+
+    # If you need to make complex sub-parts in the datasets with configurable options
+    # You can create your own builder configuration class to store attribute, inheriting from datasets.BuilderConfig
+    # BUILDER_CONFIG_CLASS = MyBuilderConfig
+
+    # You will be able to load one or the other configurations in the following list with
+    # data = datasets.load_dataset('my_dataset', 'first_domain')
+    # data = datasets.load_dataset('my_dataset', 'second_domain')
+    BUILDER_CONFIGS = [
+        datasets.BuilderConfig(
+            name="random-split-1",
+            version=VERSION,
+            description="The random-split-1-train/dev.tsv and pristine-unseen-tables.tsv",
+        ),
+        datasets.BuilderConfig(
+            name="random-split-2",
+            version=VERSION,
+            description="The random-split-2-train/dev.tsv and pristine-unseen-tables.tsv",
+        ),
+        datasets.BuilderConfig(
+            name="random-split-3",
+            version=VERSION,
+            description="The random-split-3-train/dev.tsv and pristine-unseen-tables.tsv",
+        ),
+        datasets.BuilderConfig(
+            name="random-split-4",
+            version=VERSION,
+            description="The random-split-4-train/dev.tsv and pristine-unseen-tables.tsv",
+        ),
+        datasets.BuilderConfig(
+            name="random-split-5",
+            version=VERSION,
+            description="The random-split-5-train/dev.tsv and pristine-unseen-tables.tsv",
+        ),
+    ]
+
+    DEFAULT_CONFIG_NAME = (
+        "random-split-1"  # It's not mandatory to have a default configuration. Just use one if it make sense.
+    )
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "id": datasets.Value("string"),
+                "question": datasets.Value("string"),
+                "answers": datasets.features.Sequence(datasets.Value("string")),
+                "table": {
+                    "header": datasets.features.Sequence(datasets.Value("string")),
+                    "rows": datasets.features.Sequence(datasets.features.Sequence(datasets.Value("string"))),
+                    "name": datasets.Value("string"),
+                },
+            }
+        )
+        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=features,  # Here we define them above because they are different between the two configurations
+            # If there's a common (input, target) tuple from the features, uncomment supervised_keys line below and
+            # specify them. They'll be used if as_supervised=True in builder.as_dataset.
+            # supervised_keys=("sentence", "label"),
+            # Homepage of the dataset for documentation
+            homepage=_HOMEPAGE,
+            # License for the dataset if available
+            license=_LICENSE,
+            # Citation for the dataset
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        train_file = "{}-train.tsv".format(self.config.name)
+        dev_file = "{}-dev.tsv".format(self.config.name)
+        test_file = "pristine-unseen-tables.tsv"
+        # dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLS
+        urls = _DATA_URL
+        root_dir = os.path.join(dl_manager.download_and_extract(urls), "WikiTableQuestions")
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                # These kwargs will be passed to _generate_examples
+                gen_kwargs={"main_filepath": os.path.join(root_dir, "data", train_file), "root_dir": root_dir},
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                # These kwargs will be passed to _generate_examples
+                gen_kwargs={"main_filepath": os.path.join(root_dir, "data", test_file), "root_dir": root_dir},
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                # These kwargs will be passed to _generate_examples
+                gen_kwargs={"main_filepath": os.path.join(root_dir, "data", dev_file), "root_dir": root_dir},
+            ),
+        ]
+
+    def _read_table_from_file(self, table_name: str, root_dir: str):
+        def _extract_table_content(_line: str):
+            _vals = [_.replace("\n", " ").strip() for _ in _line.strip("\n").split("\t")]
+            return _vals
+
+        rows = []
+        # assert ".csv" in _wtq_table_name
+        # use the normalized table file
+        table_name = table_name.replace(".csv", ".tsv")
+        with open(os.path.join(root_dir, table_name), "r", encoding="utf8") as table_f:
+            table_lines = table_f.readlines()
+            # the first line is header
+            header = _extract_table_content(table_lines[0])
+            for line in table_lines[1:]:
+                rows.append(_extract_table_content(line))
+        return {"header": header, "rows": rows, "name": table_name}
+
+    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
+    def _generate_examples(self, main_filepath, root_dir):
+        # The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
+        with open(main_filepath, encoding="utf-8") as f:
+            # skip the first line since it is the tsv header
+            next(f)
+            for idx, line in enumerate(f):
+                example_id, question, table_name, answer = line.strip("\n").split("\t")
+                answer = answer.split("|")
+                # must contain rows and header keys
+                table_content = self._read_table_from_file(table_name, root_dir)
+
+                yield idx, {"id": example_id, "question": question, "answers": answer, "table": table_content}