CrowdSpeech.py

# coding=utf-8
# Copyright 2021 YANDEX LLC.
#
# 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.

# Lint as: python3
"""CrowdSpeech:  Benchmark Dataset for Crowdsourced Audio Transcription."""


import json

import datasets
from datasets.tasks import Summarization


logger = datasets.logging.get_logger(__name__)

_DESCRIPTION = """\
CrowdSpeech is a publicly available large-scale dataset of crowdsourced audio transcriptions. \
It contains annotations for more than 50 hours of English speech transcriptions from more \
than 1,000 crowd workers.
"""

_URL = "https://raw.githubusercontent.com/pilot7747/VoxDIY/main/data/huggingface/"
_URLS = {
    "train-clean": _URL + "train-clean.json",
    "dev-clean": _URL + "dev-clean.json",
    "dev-other": _URL + "dev-other.json",
    "test-clean": _URL + "test-clean.json",
    "test-other": _URL + "test-other.json",
}


class CrowdSpeechConfig(datasets.BuilderConfig):
    """BuilderConfig for CrowdSpeech."""

    def __init__(self, **kwargs):
        """BuilderConfig for CrowdSpeech.

        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(CrowdSpeechConfig, self).__init__(**kwargs)


class CrowdSpeech(datasets.GeneratorBasedBuilder):
    """CrowdSpeech:  Benchmark Dataset for Crowdsourced Audio Transcription."""

    BUILDER_CONFIGS = [
        CrowdSpeechConfig(
            name="plain_text",
            version=datasets.Version("1.0.0", ""),
            description="Plain text",
        ),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "task": datasets.Value("string"),
                    "transcriptions": datasets.Value("string"),
                    "performers": datasets.Value("string"),
                    "gt": datasets.Value("string"),
                }
            ),
            supervised_keys=None,
            homepage="https://github.com/pilot7747/VoxDIY/",
            # citation=_CITATION,
            task_templates=[
                Summarization(
                    text_column="transcriptions", summary_column="gt"
                )
            ],
        )

    def _split_generators(self, dl_manager):
        downloaded_files = dl_manager.download_and_extract(_URLS)

        return [
            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train-clean"]}),
            datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files["test-clean"]}),
            datasets.SplitGenerator(name='test.other', gen_kwargs={"filepath": downloaded_files["test-other"]}),
            datasets.SplitGenerator(name='dev.clean', gen_kwargs={"filepath": downloaded_files["dev-clean"]}),
            datasets.SplitGenerator(name='dev.other', gen_kwargs={"filepath": downloaded_files["dev-clean"]}),
        ]

    def _generate_examples(self, filepath):
        """This function returns the examples in the raw (text) form."""
        logger.info("generating examples from = %s", filepath)
        with open(filepath, encoding="utf-8") as f:
            crowdspeech = json.load(f)
            for audio in crowdspeech["data"]:
                task = audio.get("task", "")
                transcriptions = audio.get("transcriptions", "")
                performers = audio.get("performers", "")
                gt = audio.get("gt", "")

                yield task, {
                    "task": task,
                    "transcriptions": transcriptions,
                    "performers": performers,
                    "gt": gt,
                }