Dataset Viewer
The dataset viewer is not available for this subset.
The default config contains 50 while it should generally contain 3 splits maximum (train/validation/test). If the splits train_0, train_1, train_10, train_11, train_12... are not used to differentiate between training and evaluation, please consider defining configs of this dataset instead. You can find how to define configs instead of splits here: https://huggingface.co/docs/hub/datasets-data-files-configuration

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.

Dataset Audio Russian Backup

This is a backup dataset with Russian audio data, split into train_0 to train_49 for tasks like text-to-speech, speech recognition, and speaker identification.

Features

  • text: Audio transcription (string).
    • speaker_name: Speaker identifier (string).
  • audio: Audio file.

Usage

Load the dataset like this:

from datasets import load_dataset
dataset = load_dataset("kijjjj/audio_data_russian_backup", split="train_0")  # Or any train_X split
print(dataset[0])
Downloads last month
12