megengine.module.init.msra_normal_

msra_normal_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu')[源代码]

Fills tensor wilth random values sampled from \(\mathcal{N}(0, \text{std}^2)\) where

\[\text{std} = \sqrt{\frac{2}{(1 + a^2) \times \text{fan_in}}}\]

参考 “Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification” 获取详细说明

参数
  • tensor (Tensor) – 待初始化的n维张量

  • a (float) – 用于计算leaky_relu增益的可选参数。请参阅: calculate_gain 了解详情。

  • mode (str) – fan_in” or “fan_out”, used to calculate \(gain\), the scaling factor for \(gain\). See calculate_fan_in_and_fan_out for details.

  • nonlinearity (str) – 用于计算 \(gain\) 的非线性函数的名称。请参阅 calculate_gain 了解详情。

返回类型

None