1.jupyter 实现二层卷积神经网络
2.实现步骤
设置训练数据;
设置model,loss,optimizer;
进行训练迭代(1.forward;2.loss计算;3.backward;4.参数更新)
3.注意函数的写法及传递的参数
torch.nn.Sequential()
torch.nn.Linear(Cin,Cout,bias=False)
torch.nn.ReLU()
torch.nn.MSELoss(reduction='sum')
torch.optim.Adam(model.parameters().lr=learning_rate)
torch.nn.init.normal_(model[0].weight)
y_pred=model(x)
loss=loss_fn(y_pred,y)
optimizer.zero_grad()
loss.backward()
optimizer.step()