Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/inspect.py", line 299, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/packaged_modules/folder_based_builder/folder_based_builder.py", line 229, in _split_generators
                  raise ValueError(
              ValueError: `file_name` or `*_file_name` must be present as dictionary key (with type string) in metadata files
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 65, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/inspect.py", line 353, in get_dataset_split_names
                  info = get_dataset_config_info(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/inspect.py", line 304, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

CSS100-LJSpeech (Japanese / Meian)

css100-ljspeech は、Park et al. が公開した CSS10 日本語コーパス(明暗)を、LJ Speech 互換フォーマット (id|text & wavs/*.wav) へ変換した派生データセットです。

データ概要

項目
話者 1 (ekzemplaro)
音声数 6,841
合計時間 約 15 時間
サンプリングレート 22,050 Hz
テキスト言語 日本語
フォーマット ``id

ファイル構成

css100-ljspeech/
├── metadata.csv   # 2 列 (id|text)
└── wavs/
    ├── meian_0000.wav
    ├── meian_0001.wav
    └── ...

使用例 (🤗 Datasets)

from datasets import load_dataset, Audio

ds = load_dataset("your-username/css100-ljspeech", split="train", streaming=True)
# ``id`` と ``text`` の 2 カラムがあり、``audio`` は "wavs/{id}.wav" を自動デコード
sample = next(iter(ds.cast_column("audio", Audio(sampling_rate=22050))))
print(sample["id"], sample["text"])
print(sample["audio"].keys())  # array, sampling_rate など

CLI

huggingface-cli download ayousanz/css10-ja-ljspeech --repo-type dataset --local-dir path/to/download/ --local-dir-use-symlinks False

ライセンス

元データセット CSS10Apache License 2.0 の下で提供されています。本派生物 (変換スクリプト・フォーマット変更) も同じ Apache-2.0 ライセンスで公開します。元オーディオは LibriVox を出典とし、パブリックドメイン音声を含みますが、再配布にあたり Apache-2.0 の条項に従ってください。

引用

@inproceedings{park2019css10,
  title={CSS10: A Collection of Single Speaker Speech Datasets for 10 Languages},
  author={Park, Kyubyong and Mulc, Thomas},
  booktitle={Proc. Interspeech},
  year={2019}
}
Downloads last month
312