道
V1
2022/01/06阅读:59主题:默认主题
pytorch中Parameter()介绍
用法介绍
pytorch中的Parameter函数可以对某个张量进行参数化。它可以将不可训练的张量转化为可训练的参数类型,同时将转化后的张量绑定到模型可训练参数的列表中,当更新模型的参数时一并将其更新。
torch.nn.parameter.Parameter
data (Tensor):表示需要参数化的张量 requires_grad (bool, optional):表示是否该张量是否需要梯度,默认值为True
代码介绍
pytorch中的Parameter函数具体的代码示例如下所示
import torch
import torch.nn as nn
class NeuralNetwork(nn.Module):
def __init__(self, input_dim, output_dim):
super(NeuralNetwork, self).__init__()
self.linear = nn.Linear(input_dim, output_dim)
self.linear.weight = torch.nn.Parameter(torch.zeros(input_dim, output_dim))
self.linear.bias = torch.nn.Parameter(torch.ones(output_dim))
def forward(self, input_array):
output = self.linear(input_array)
return output
if __name__ == '__main__':
net = NeuralNetwork(4, 6)
for param in net.parameters():
print(param)
代码的结果如下所示:
当神经网络的参数不是用Parameter函数参数化直接赋值给权重参数时,则会报错,具体的程序
import torch
import torch.nn as nn
class NeuralNetwork(nn.Module):
def __init__(self, input_dim, output_dim):
super(NeuralNetwork, self).__init__()
self.linear = nn.Linear(input_dim, output_dim)
self.linear.weight = torch.zeros(input_dim, output_dim)
self.linear.bias = torch.ones(output_dim)
def forward(self, input_array):
output = self.linear(input_array)
return output
if __name__ == '__main__':
net = NeuralNetwork(4, 6)
for param in net.parameters():
print(param)
代码运行报错结果如下所示:
作者介绍
道
V1