Pytorch 的参数初始化 - torch.nn.init
给定非线性函数的推荐增益值(gain value):
nonlinearity 非线性函数 | gain 增益 |
Linear / Identity | 1 |
Conv{1,2,3}D | 1 |
Sigmoid | 1 |
Tanh | $\frac{5}{3}$ |
ReLU | $\sqrt{2}$ |
Leaky Relu | $\sqrt{\frac{2}{1 + \text{negative_slope}^2}}$ |
Gain is a proportional value that shows the relationship between the magnitude of the input to the magnitude of the output signal at steady state. Many systems contain a method by which the gain can be altered, providing more or less "power" to the system.
-- From wiki.
1. torch.nn.init.calculate_gain
torch.nn.init.calculate_gain(nonlinearity, param=None)
- nonlinearlity - 非线性函数名
- param - 非线性函数的可选参数
如:
import torch.nn as nn
gain = nn.init.calculate_gain('leaky_relu')
>>> 1.414...
2. torch.nn.init 初始化函数
import torch
import torch.nn as nn
w = torch.empty(2, 3)
# 1. 均匀分布 - u(a,b)
# torch.nn.init.uniform_(tensor, a=0, b=1)
nn.init.uniform_(w)
# tensor([[ 0.0578, 0.3402, 0.5034],
# [ 0.7865, 0.7280, 0.6269]])
# 2. 正态分布 - N(mean, std)
# torch.nn.init.normal_(tensor, mean=0, std=1)
nn.init.normal_(w)
# tensor([[ 0.3326, 0.0171, -0.6745],
# [ 0.1669, 0.1747, 0.0472]])
# 3. 常数 - 固定值 val
# torch.nn.init.constant_(tensor, val)
nn.init.constant_(w, 0.3)
# tensor([[ 0.3000, 0.3000, 0.3000],
# [ 0.3000, 0.3000, 0.3000]])
# 4. 对角线为 1,其它为 0
# torch.nn.init.eye_(tensor)
nn.init.eye_(w)
# tensor([[ 1., 0., 0.],
# [ 0., 1., 0.]])
# 5. Dirac delta 函数初始化,仅适用于 {3, 4, 5}-维的 torch.Tensor
# torch.nn.init.dirac_(tensor)
w1 = torch.empty(3, 16, 5, 5)
nn.init.dirac_(w1)
# 6. xavier_uniform 初始化
# torch.nn.init.xavier_uniform_(tensor, gain=1)
# From - Understanding the difficulty of training deep feedforward neural networks - Bengio 2010
nn.init.xavier_uniform_(w, gain=nn.init.calculate_gain('relu'))
# tensor([[ 1.3374, 0.7932, -0.0891],
# [-1.3363, -0.0206, -0.9346]])
# 7. xavier_normal 初始化
# torch.nn.init.xavier_normal_(tensor, gain=1)
nn.init.xavier_normal_(w)
# tensor([[-0.1777, 0.6740, 0.1139],
# [ 0.3018, -0.2443, 0.6824]])
# 8. kaiming_uniform 初始化
# From - Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification - HeKaiming 2015
# torch.nn.init.kaiming_uniform_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu')
nn.init.kaiming_uniform_(w, mode='fan_in', nonlinearity='relu')
# tensor([[ 0.6426, -0.9582, -1.1783],
# [-0.0515, -0.4975, 1.3237]])
# 9. kaiming_normal 初始化
# torch.nn.init.kaiming_normal_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu')
nn.init.kaiming_normal_(w, mode='fan_out', nonlinearity='relu')
# tensor([[ 0.2530, -0.4382, 1.5995],
# [ 0.0544, 1.6392, -2.0752]])
# 10. 正交矩阵 - (semi)orthogonal matrix
# From - Exact solutions to the nonlinear dynamics of learning in deep linear neural networks - Saxe 2013
# torch.nn.init.orthogonal_(tensor, gain=1)
nn.init.orthogonal_(w)
# tensor([[ 0.5786, -0.5642, -0.5890],
# [-0.7517, -0.0886, -0.6536]])
# 11. 稀疏矩阵 - sparse matrix
# 非零元素采用正态分布 N(0, 0.01) 初始化.
# From - Deep learning via Hessian-free optimization - Martens 2010
# torch.nn.init.sparse_(tensor, sparsity, std=0.01)
nn.init.sparse_(w, sparsity=0.1)
# tensor(1.00000e-03 *
# [[-0.3382, 1.9501, -1.7761],
# [ 0.0000, 0.0000, 0.0000]])