VIST.py

# 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.
# TODO: Address all TODOs and remove all explanatory comments
"""TODO: Add a description here."""


import json
import os
from huggingface_hub import hf_hub_download
import datasets


# TODO: Add BibTeX citation
# Find for instance the citation on arxiv or on the dataset repo/website
_CITATION = """\
@InProceedings{huggingface:dataset,
title = {A great new dataset},
author={huggingface, Inc.
},
year={2020}
}
"""

# TODO: Add description of the dataset here
# You can copy an official description
_DESCRIPTION = """\
This new dataset is designed to solve this great NLP task and is crafted with a lot of care.
"""

# TODO: Add a link to an official homepage for the dataset here
_HOMEPAGE = "http://visionandlanguage.net/VIST/dataset.html"

# TODO: Add the licence for the dataset here if you can find it
_LICENSE = ""

_DII_WORKER_IDS = hf_hub_download(repo_type="dataset", repo_id="society-ethics/VIST", filename="data/dii.worker_ids.csv", use_auth_token=True)

_URLS = {
    "DII": {
        "train": "https://huggingface.co/datasets/NimaBoscarino/VIST/resolve/main/data/train.dii.jsonl.zip",
        "test": "https://huggingface.co/datasets/NimaBoscarino/VIST/resolve/main/data/test.dii.jsonl.zip",
        "val": "https://huggingface.co/datasets/NimaBoscarino/VIST/resolve/main/data/val.dii.jsonl.zip",
    },
    "SIS": "http://visionandlanguage.net/VIST/json_files/story-in-sequence/SIS-with-labels.tar.gz",
}


# TODO: Name of the dataset usually matches the script name with CamelCase instead of snake_case
class VIST(datasets.GeneratorBasedBuilder):
    """TODO: Short description of my dataset."""

    VERSION = datasets.Version("1.1.0")

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="DII", version=VERSION, description=""),
        datasets.BuilderConfig(name="SIS", version=VERSION, description=""),
    ]

    def _info(self):
        features = None
        if self.config.name == "DII":
            features = datasets.Features({
                'description': datasets.Value("string"),
                'title': datasets.Value("string"),
                'farm': datasets.ClassLabel(num_classes=10),  # Actually 9, but datasets complains for some reason?
                'date_update': datasets.Value("timestamp[s]"),
                'primary': datasets.Value("int32"),
                'server': datasets.Value("int16"),
                'date_create': datasets.Value("timestamp[s]"),
                'photos': datasets.Value("int16"),
                'secret': datasets.Value("string"),
                'owner': datasets.Value("string"),
                'vist_label': datasets.Value("string"),
                'id': datasets.Value("int64"),
                "images": datasets.Sequence({
                    'datetaken': datasets.Value("date64"),
                    'license': datasets.ClassLabel(num_classes=7),
                    'image_title': datasets.Value("string"),
                    'longitude': datasets.Value("float64"),
                    'url': datasets.Image(decode=False),
                    'image_secret': datasets.Value("string"),
                    'media': datasets.ClassLabel(num_classes=2, names=["photo", "video"]),
                    'latitude': datasets.Value("float64"),
                    'image_id': datasets.Value("int64"),
                    'tags': [datasets.Value("string")],
                    'image_farm': datasets.ClassLabel(names=["1", "2", "6", "7"]),  # From exploring the data
                    'image_server': datasets.Value("int16"),
                    "annotations": datasets.Sequence({
                        'original_text': datasets.Value("string"),
                        'photo_order_in_story': datasets.Value("int8"),
                        'worker_id': datasets.ClassLabel(names_file=_DII_WORKER_IDS),
                        'text': datasets.Value("string"),
                    })
                })
            })
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            homepage=_HOMEPAGE,
            license=_LICENSE,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        urls = _URLS[self.config.name]
        data_dirs = dl_manager.download_and_extract(urls)

        for split in data_dirs:
            archive_path = data_dirs[split]
            if archive_path.endswith(".zip") or os.path.isdir(archive_path):
                data_dirs[split] = os.path.join(archive_path, os.listdir(archive_path)[0])

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "filepath": data_dirs["train"],
                    "split": "train",
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                gen_kwargs={
                    "filepath": data_dirs["val"],
                    "split": "validation",
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={
                    "filepath": data_dirs["test"],
                    "split": "test"
                },
            ),
        ]

    def _generate_examples(self, filepath, split):
        with open(filepath, encoding="utf-8") as f:
            for key, row in enumerate(f):
                data = json.loads(row)
                yield key, data