espnet2.gan_codec.hificodec.module.init_weights

source

espnet2.gan_codec.hificodec.module.init_weights(m, mean=0.0, std=0.01)

Initialize weights for convolutional layers.

This function applies a normal distribution to the weights of convolutional layers in a given module. It sets the weights to have a mean of mean and a standard deviation of std.

• Parameters:
  • m (torch.nn.Module) – The module whose weights are to be initialized.
  • mean (float , optional) – The mean of the normal distribution. Defaults to 0.0.
  • std (float , optional) – The standard deviation of the normal distribution. Defaults to 0.01.

NOTE

This function specifically checks if the module is of a convolutional type (i.e., classes that contain “Conv” in their name) before applying the weight initialization.

Examples

>>> model = nn.Sequential(
...     nn.Conv2d(1, 20, 5),
...     nn.Conv2d(20, 64, 5)
... )
>>> model.apply(init_weights)