---
task_categories:
- audio-to-audio
tags:
- music
pretty_name: YouTubeBigBand
size_categories:
- n<1K
---
# YouTubeBigBand Dataset

Inspired by the [YouTubeMix](https://huggingface.co/datasets/krandiash/youtubemix) dataset.
<br/>
<br/>
*Source*: [https://www.youtube.com/watch?v=I4KAKqF4mjE](https://www.youtube.com/watch?v=I4KAKqF4mjE) - a 2 hour long mix of jazz tracks played by a big band.
<br/>
<br/>
Used for pre-training a [SaShiMi model (see citation)](https://arxiv.org/abs/2202.09729) as part of Tel Aviv University Deep Learning Workshop 2024 Semester B (see [project repository fork](https://github.com/galbezalel/s4-dl-workshop/)).

We include two versions of the dataset:
- `youtubebigband.zip` is a zip file containing 129 1-minute audio clips (re)sampled at 16kHz. These were generated by splitting the original audio track. 
- `raw_bigband.wav` is the raw audio track from the YouTube video, sampled at 44.1kHz.

```
@article{goel2022sashimi,
  title={It's Raw! Audio Generation with State-Space Models},
  author={Goel, Karan and Gu, Albert and Donahue, Chris and R\'{e}, Christopher},
  journal={arXiv preprint arXiv:2202.09729},
  year={2022}
}

@misc{deepsound,
  author = {DeepSound},
  title = {SampleRNN},
  year = {2017},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/deepsound-project/samplernn-pytorch}},
}
```