net_to_equal_lr¶

class torchelie.nn.utils.net_to_equal_lr(net: torch.nn.modules.module.Module, leak: float = 0.0, mode: str = 'fan_in')¶

Set all Conv2d, ConvTransposed2d and Linear of net to equalized learning rate, initialized with torchelie.utils.kaiming() and dynamic=True.