CNN-Renew

Pytorch/Project/ez0/MyFirstModule.py

@@ -0,0 +1,49 @@
+import torch
+import torch.nn as nn
+
+ez = nn.Sequential(
+    nn.Linear(2, 8),
+    nn.ReLU(),
+    nn.Linear(8, 8),
+    nn.ReLU(),
+    nn.Linear(8, 1)
+)
+# y = w1 @ x + w2 @ x^2 + b
+# 两个特征，第一个是x，第二个是x^2
+criterion = nn.MSELoss()
+
+x = torch.linspace(-10, 10, 10000).unsqueeze(1)
+X = torch.cat([x, x**2], dim=1)
+y = x**2 + 2 * x + 1
+
+inx = torch.tensor([[2, 4], [1, 1], [3, 9], [4, 16]], dtype=torch.float32)
+target = torch.tensor([[9], [4], [16], [25]], dtype=torch.float32)
+
+lr = 0.00001
+for i in range(10000):
+    #向前传播
+    out = ez(X)
+    loss = criterion(out, y)
+
+    #反向传播
+    loss.backward()
+
+    with torch.no_grad():
+        # 3. 手动更新参数
+        with torch.no_grad():  # 禁止 autograd 追踪
+            for param in ez.parameters():
+                param -= lr * param.grad  # 梯度下降更新参数
+
+    #梯度清零
+    ez.zero_grad()
+
+    if i % 1000 == 0:
+        print(f"Epoch {i}: loss={loss.item():.4f}")
+
+import matplotlib.pyplot as plt
+
+y_pred = ez(X).detach()
+plt.scatter(x.numpy(), x.numpy()**2 + 2*x.numpy() + 1, label='real y = x^2 + 2x + 1')
+plt.plot(x.numpy(), y_pred.numpy(), color='r', label='Predicted y = x^2 + 2x + 1')
+plt.legend()
+plt.show()