|
--- |
|
license: mit |
|
thumbnail: "https://iscale.iheart.com/catalog/album/46707655" |
|
tags: |
|
- audio |
|
- music |
|
- generation |
|
- tensorflow |
|
--- |
|
|
|
# Musika Model: musika-grateful-dead-barton-hall |
|
## Model provided by: benwakefield |
|
|
|
Pretrained model for the [Musika system](https://github.com/marcoppasini/musika) for fast infinite waveform music generation. |
|
Introduced in [this paper](https://arxiv.org/abs/2208.08706). |
|
|
|
Trained on the [Cornell 5/8/77](https://en.wikipedia.org/wiki/Cornell_5/8/77) show performed by the Grateful Dead. |
|
|
|
## How to use |
|
|
|
You can generate music from this model using the notebook available [here](https://colab.research.google.com/drive/1HJWliBXPi-Xlx3gY8cjFI5-xaZgrTD7r). |
|
|
|
### Model description |
|
|
|
This pretrained GAN system consists of a ResNet-style generator and discriminator. During training, stability is controlled by adapting the strength of gradient penalty regularization on-the-fly. The gradient penalty weighting term is contained in *switch.npy*. The generator is conditioned on a latent coordinate system to produce samples of arbitrary length. The latent representations produced by the generator are then passed to a decoder which converts them into waveform audio. |
|
The generator has a context window of about 12 seconds of audio. |
|
|
