task2.py

Tasks/daily tasks/Jamcey_V_P/task2.py

@@ -0,0 +1,34 @@
+import torch
+import torch.nn as nn
+
+class neural_network(nn.Module):
+  def __init__(self, ):
+    super(neural_network, self).__init__()
+
+    self.input_layer = 4                  # Number of input units
+    self.hidden_layer1 = 5                # Number of hidden units
+    self.hidden_layer2 = 3                # Number of hidden units  
+    self.output_layer = 1                # Number of output units 
+
+    # Weights 
+    W1 = torch.randn(self.input_layer, self.hidden_layer1)  
+    W2 = torch.randn(self.hidden_layer1, self.hidden_layer2) 
+    W3 = torch.randn(self.hidden_layer2, self.output_layer) 
+
+
+    #  bias 
+    B1 = torch.randn((1, self.hidden_layer1)) 
+    B2 = torch.randn((1,self.hidden_layer2))
+    B3 = torch.randn((1,self.output_layer))
+
+    def forward(self, X):
+      z1 = torch.mm(X, w1) + b1
+      Relu = nn.ReLU()
+      a1 = Relu(z1)
+      z2 = torch.mm(X, w2) + b2
+      Relu = nn.ReLU()
+      a2 = Relu(z2)
+      z3 = torch.mm(X, w3) + b3
+      Sigmoid = nn.Sigmoid()
+      Result = Sigmoid(z3)
+      return Result