关于Resnet与vgg集成训练精度提不上去问题

各位大神，为什么我的模型训练后卡在了72左右上下浮动，val的loss值也没有降下来，是因为模型出问题了吗，我觉得模型应该不算太简单的， 引用了resnet预训练模型，训练集一共1800张左右，也不算太小，求解

下面附上运行图

下面附上模型部分代码

import torch
import torch.nn as nn
import torchvision.models as models

class BJWL2(nn.Module):
    def __init__(self, num_classes=10):
        super(BJWL2, self).__init__()
        self.features_vgg = nn.Sequential(
            nn.Conv2d(3, 64, kernel_size=3, padding=1),
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=2, stride=2),

            nn.Conv2d(64, 128, kernel_size=3, padding=1),
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=2, stride=2),

            nn.Conv2d(128, 256, kernel_size=3, padding=1),
            nn.ReLU(inplace=True),
            nn.Conv2d(256, 256, kernel_size=3, padding=1),
            nn.ReLU(inplace=True),
            nn.MaxPool2d(kernel_size=2, stride=2),

            nn.Conv2d(256, 512, kernel_size=3, padding=1),  # 添加一个卷积层
            nn.ReLU(inplace=True),
        )
        self.avgpool_vgg = nn.AdaptiveAvgPool2d((5, 5))
        self.classifier_vgg = nn.Sequential(
            nn.Linear(512 * 5 * 5, 2048),  # 增加全连接层的维度
            nn.ReLU(True),
            nn.Dropout(),
            nn.Linear(2048, 1024),
            nn.ReLU(True),
            nn.Dropout(),
            nn.Linear(1024, num_classes),
        )

        self.resnet = models.resnet18(pretrained=True)
        num_ftrs = self.resnet.fc.in_features
        self.resnet.fc = nn.Linear(num_ftrs, num_classes)

    def forward(self, x):
        features_vgg = self.features_vgg(x)
        x_vgg = self.avgpool_vgg(features_vgg)
        x_vgg = torch.flatten(x_vgg, 1)
        x_vgg = self.classifier_vgg(x_vgg)

        x_resnet = self.resnet(x)

        # 合并 VGG 和 ResNet 的输出
        combined_output = x_vgg + x_resnet

        return combined_output

# 创建一个结合了 VGG 和 ResNet 的模型实例
vgg = BJWL2(num_classes=10)
print(vgg)