File size: 496 Bytes
9a78ded | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 | from transformers import PretrainedConfig
class SoundStreamConfig(PretrainedConfig):
model_type = "soundstream"
def __init__(
self,
channels: int = 32,
latent_dim: int = 512,
codebook_size: int = 1024,
num_quantizers: int = 8,
**kwargs,
):
super().__init__(**kwargs)
self.channels = channels
self.latent_dim = latent_dim
self.codebook_size = codebook_size
self.num_quantizers = num_quantizers
|