You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
46 lines
1.5 KiB
46 lines
1.5 KiB
1 year ago
|
import torch
|
||
|
|
|
||
|
|
torch.manual_seed(2023)
|
||
|
|
|
||
|
|
|
||
|
|
def activation_func(x):
|
||
|
|
#TODO Implement one of these following activation function: sigmoid, tanh, ReLU, leaky ReLU
|
||
|
|
epsilon = 0.01 # Only use this variable if you choose Leaky ReLU
|
||
|
|
result = None
|
||
|
|
return result
|
||
|
|
|
||
|
|
def softmax(x):
|
||
|
|
# TODO Implement softmax function here
|
||
|
|
result = None
|
||
|
|
return result
|
||
|
|
|
||
|
|
|
||
|
|
# Define the size of each layer in the network
|
||
|
|
num_input = 784 # Number of node in input layer (28x28)
|
||
|
|
num_hidden_1 = 128 # Number of nodes in hidden layer 1
|
||
|
|
num_hidden_2 = 256 # Number of nodes in hidden layer 2
|
||
|
|
num_hidden_3 = 128 # Number of nodes in hidden layer 3
|
||
|
|
num_classes = 10 # Number of nodes in output layer
|
||
|
|
|
||
|
|
# Random input
|
||
|
|
input_data = torch.randn((1, num_input))
|
||
|
|
# Weights for inputs to hidden layer 1
|
||
|
|
W1 = torch.randn(num_input, num_hidden_1)
|
||
|
|
# Weights for hidden layer 1 to hidden layer 2
|
||
|
|
W2 = torch.randn(num_hidden_1, num_hidden_2)
|
||
|
|
# Weights for hidden layer 2 to hidden layer 3
|
||
|
|
W3 = torch.randn(num_hidden_2, num_hidden_3)
|
||
|
|
# Weights for hidden layer 3 to output layer
|
||
|
|
W4 = torch.randn(num_hidden_3, num_classes)
|
||
|
|
|
||
|
|
# and bias terms for hidden and output layers
|
||
|
|
B1 = torch.randn((1, num_hidden_1))
|
||
|
|
B2 = torch.randn((1, num_hidden_2))
|
||
|
|
B3 = torch.randn((1, num_hidden_3))
|
||
|
|
B4 = torch.randn((1, num_classes))
|
||
|
|
|
||
|
|
#TODO Calculate forward pass of the network here. Result should have the shape of [1,10]
|
||
|
|
# Dont forget to check if sum of result = 1.0
|
||
|
|
result = None
|
||
|
|
print(result)
|