G_Tlover的博客

函数列表torch.flatten()函数nn.nn.Linear()函数torch.flatten()函数A = torch.tensor([[[1,2,3,4],[5,6,7,8],[9,10,11,12]],[[13,14,15,16],[17,18,19,20],[21,22,23,24]]])out1 = torch.flatten(A)out2 = torch.flatten(A,0)out3 = torch.flatten(A,1)print("默认 = ",out1)print

文章目录1. DarkNet网络结构1.1 卷积单元----Convolutional1.2 Residual二、Pytorch实现三、Tensorboard绘制网络结构图1. DarkNet网络结构Yolov3、Yolov4的backbone采用了CSP_DarkNet_53网络,其激活函数采用Mish。本文章复现DarkNet_53激活函数采用LeakyReLu，两者结构基本一致。1.1 卷积单元----ConvolutionalConvolutional卷积单元由Conv+BN+LeakyRe

pytorch框架下两种模型保存及加载方式方法1以字典（key-value）保存模型的参数torch.save(vgg16.state_dict(), "路径")vgg16.load_state_dict(torch.load("路径"))方法2保存模型参数及模型结构torch.save(vgg16, "路径h")model = torch.load("路径")...