torchaudio.datasets — Torchaudio 2.7.0 documentation (original) (raw)
All datasets are subclasses of torch.utils.data.Datasetand have __getitem__ and __len__ methods implemented.
Hence, they can all be passed to a torch.utils.data.DataLoaderwhich can load multiple samples parallelly using torch.multiprocessing workers. For example:
yesno_data = torchaudio.datasets.YESNO('.', download=True) data_loader = torch.utils.data.DataLoader( yesno_data, batch_size=1, shuffle=True, num_workers=args.nThreads)
| CMUARCTIC | CMU ARCTIC [Kominek et al., 2003] dataset. |
|---|---|
| CMUDict | CMU Pronouncing Dictionary [Weide, 1998] (CMUDict) dataset. |
| COMMONVOICE | CommonVoice [Ardila et al., 2020] dataset. |
| DR_VCTK | Device Recorded VCTK (Small subset version) [Sarfjoo and Yamagishi, 2018] dataset. |
| FluentSpeechCommands | Fluent Speech Commands [Lugosch et al., 2019] dataset |
| GTZAN | GTZAN [Tzanetakis et al., 2001] dataset. |
| IEMOCAP | IEMOCAP [Busso et al., 2008] dataset. |
| LibriMix | LibriMix [Cosentino et al., 2020] dataset. |
| LIBRISPEECH | LibriSpeech [Panayotov et al., 2015] dataset. |
| LibriLightLimited | Subset of Libri-light [Kahn et al., 2020] dataset, which was used in HuBERT [Hsu et al., 2021] for supervised fine-tuning. |
| LIBRITTS | LibriTTS [Zen et al., 2019] dataset. |
| LJSPEECH | LJSpeech-1.1 [Ito and Johnson, 2017] dataset. |
| MUSDB_HQ | MUSDB_HQ [Rafii et al., 2019] dataset. |
| QUESST14 | QUESST14 [Miro et al., 2015] dataset. |
| Snips | Snips [Coucke et al., 2018] dataset. |
| SPEECHCOMMANDS | Speech Commands [Warden, 2018] dataset. |
| TEDLIUM | Tedlium [Rousseau et al., 2012] dataset (releases 1,2 and 3). |
| VCTK_092 | VCTK 0.92 [Yamagishi et al., 2019] dataset |
| VoxCeleb1Identification | VoxCeleb1 [Nagrani et al., 2017] dataset for speaker identification task. |
| VoxCeleb1Verification | VoxCeleb1 [Nagrani et al., 2017] dataset for speaker verification task. |
| YESNO | YesNo [YesNo, n.d.] dataset. |