1.用法:
G_skeleton, D_skeleton, start_epoch = self.build_models()
optimizerG_skeleton, optimizerD_skeleton = self.define_optimizers(G_skeleton, D_skeleton)
criterion = nn.BCELoss()
其中G_skeleton
和 D_skeleton
是我們用到的模型。使用以下代碼打印參數總數:
# 打印G和D的總參數數量
print("Total number of param in Generator is ", sum(x.numel() for x in G_skeleton.parameters()))
print("Total number of param in Discriminator is ", sum(x.numel() for x in D_skeleton.parameters()))
2.解析:
my_model.parameters()
:用來返回模型中的參數
numel()
:獲取tensor中一共包含多少個元素
例:
sum()
:python內置函數,對元組或列表求和