|
--- |
|
license: mit |
|
library_name: keras |
|
tags: |
|
- audio |
|
- music |
|
- generation |
|
- tensorflow |
|
--- |
|
|
|
# Musika Misc Small Model |
|
|
|
Pretrained Misc Small GAN 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). |
|
|
|
## 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 6 seconds of audio (half the context of the full Misc model). |
|
|
|
### How to use |
|
|
|
This pretrained Misc Small GAN system is automatically downloaded at the first execution of the system. Try Musika [here](https://github.com/marcoppasini/musika)! |
|
|
|
|
|
## Training data |
|
|
|
The Misc Small GAN system was trained on the SXSW music dataset (17000 songs with diverse genres). |