datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
librispeech40ru_librispeech_for_speaker_separationDataset for source audio separation task based on Russian LibriSpeech (RuLS) dataset. Dataset contains 50 000 audio mixtures with 2 speakers for train part; 12500 audio mixtures for test part.
Dataset also containts metadata files with audio duration (sec), source 1 and source 2 filepaths for each audio mixture.
source: https://www.openslr.org/96/
librispeech_asr_dummy_orthographmetavoice_librispeech-long_LibriTTS
id
sentence
121_127105_000043_000004
He was handsome and bold and pleasant, offhand and gay and kind.
121_127105_000040_000000
"With this outbreak at last."
121_127105_000012_000001
He passed his hand over his eyes, made a little wincing grimace.
Speaker\Text
121_127105_000043_000004
121_127105_000040_000000
121_127105_000012_000001
7850
🎧
🎧
🎧
6313
🎧
🎧
🎧
422
🎧
🎧
🎧
2086
🎧
🎧
🎧
5895
🎧
🎧
🎧
2803
🎧
🎧
🎧
1919
🎧
🎧
🎧
6319
🎧
🎧
🎧
5536… See the full description on the dataset page: https://huggingface.co/datasets/Tony-Yeh/metavoice_librispeech-long_LibriTTS.joefox_LibriSpeech_test_noise_test_embeddings