Function xavier_normal_

imperative/python/megengine/module/init.py:199–217

r"""Fills tensor with random values sampled from :math:`\mathcal{N}(0, \text{std}^2)` where .. math:: \text{std} = \text{gain} \times \sqrt{\frac{2}{\text{fan_in} + \text{fan_out}}} Also known as Glorot initialization. Detailed information can be retrieved from `Understand

(tensor: Tensor, gain: float = 1.0)

def xavier_normal_(tensor: Tensor, gain: float = 1.0) -> None:
    r"""Fills tensor with random values sampled from
    :math:`\mathcal{N}(0, \text{std}^2)` where

    .. math::

        \text{std} = \text{gain} \times \sqrt{\frac{2}{\text{fan_in} + \text{fan_out}}}

    Also known as Glorot initialization. Detailed information can be retrieved from
    `Understanding the difficulty of training deep feedforward neural networks` -
    Glorot, X. & Bengio, Y. (2010).

    Args:
        tensor: tensor to be initialized.
        gain: scaling factor for :math:`std`.
    """
    fan_in, fan_out = calculate_fan_in_and_fan_out(tensor)
    std = gain * math.sqrt(2.0 / float(fan_in + fan_out))
    normal_(tensor, 0.0, std)


def msra_uniform_(