动手深度学习NIN提问

Question

以下是我的代码，运行后出现Boolean value of Tensor with more than one value is ambiguous的错误，请问该如何修改呢？

import time
import torch
from torch import nn,optim

import sys
sys.path.append("C:/Users/zyx20/Desktop/深度学习编程/pythonProject")
import d2lzh_pytorch as d2l
#定义NIN块
def nin_block(in_channels,out_channels,kernel_size,stride,padding):
    blk=nn.Sequential(nn.Conv2d(in_channels,out_channels,kernel_size,stride,padding),
                      nn.ReLU(),
                      nn.Conv2d(out_channels,out_channels,kernel_size=1),
                      nn.ReLU(),
                      nn.Conv2d(out_channels,out_channels,kernel_size=1),
                      nn.ReLU())
    return blk

class GlobalAvgPool2d(nn.Module):
    def __init__(self):
        super(GlobalAvgPool2d,self).__init__()
    def forward(self,x):
        return torch.nn.functional.avg_pool2d(x,kernel_size=x.size()[2:])

net=nn.Sequential(
    nin_block(1,96,kernel_size=11,stride=4,padding=0),
    nn.MaxPool2d(kernel_size=3,stride=2),
    nin_block(96,256,kernel_size=5,stride=1,padding=2),
    nn.MaxPool2d(kernel_size=3,stride=2),
    nin_block(256,384,kernel_size=3,stride=1,padding=1),
    nn.MaxPool2d(kernel_size=3,stride=2),
    nn.Dropout(0.5),
    nin_block(384,10,kernel_size=3,stride=1,padding=1),
    GlobalAvgPool2d(),
    d2l.FlattenLayer())
#训练模型
batch_size=128
train_iter,test_iter=d2l.load_data_fashion_mnist(batch_size,resize=224)
#训练
lr,num_epochs=0.002,5
optimizer=torch.optim.Adam(net.parameters(),lr=lr)
d2l.train_ch5(net,train_iter,test_iter,num_epochs,batch_size,optimizer,num_epochs)

Answer 1

tjdnbj 上午好☀️☀️☀️️
本答案参考通义千问

该错误提示Boolean value of Tensor with more than one value is ambiguous是因为你在某些操作中尝试将一个包含多个值（如张量）转换为布尔值，但布尔值只能表示真或假。在你的代码中，可能是nn.functional.avg_pool2d(x, kernel_size=x.size()[2:])这行有问题，因为kernel_size参数可能是一个形状为(H, W)的张量。

为了解决这个问题，你需要确保kernel_size是一个标量，而不是一个张量。你可以通过kernel_size=int(kernel_size)来实现。下面是修改后的代码：

class GlobalAvgPool2d(nn.Module):
    def __init__(self):
        super(GlobalAvgPool2d, self).__init__()
    def forward(self, x):
        kernel_size = int(x.size()[2:])
        return torch.nn.functional.avg_pool2d(x, kernel_size=kernel_size)

# 在定义GlobalAvgPool2d时，对kernel_size进行处理
net = nn.Sequential(
    # 其他代码保持不变
    GlobalAvgPool2d()
    # ...其他层...
)

这样，kernel_size就会被正确地转换为一个整数，避免了布尔值的混淆。