T5-Machine-Learning/main.py at main · sofiachavezb/T5-Machine-Learning · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
from src import LinearLayerModel, train, SGD, accuracy, confusion_matrix, BasicCNN
from torch.optim import Adam
import matplotlib.pyplot as plt

input_size = 784  # for Fashion MNIST
output_size = 10  # for Fashion MNIST
batch_size = 1000
epochs = 10

model = LinearLayerModel(input_size, output_size)

print("Training with custom SGD...")
custom_optimizer_sgd = SGD(model.parameters(), lr=0.1)
SGD_losses,SGD_preds, SGD_labels =train(model, custom_optimizer_sgd, batch_size, epochs)

print("Training with Adam...")
model = LinearLayerModel(input_size, output_size)
optimizer_adam = Adam(model.parameters(), lr=0.01)
Adam_losses, Adam_preds, Adam_labels = train(model, optimizer_adam, batch_size=100, epochs=10)

# Plot the losses
plt.plot(SGD_losses, label="SGD")
plt.plot(Adam_losses, label="Adam")
plt.title("Loss vs Epochs for Fashion MNIST dataset")
plt.xlabel("Epoch")
plt.ylabel("Loss")
plt.legend()
plt.show()

# Plot the confusion matrix
print(f"Accuracy with custom SGD: {accuracy(SGD_preds, SGD_labels)}")
print(f"Accuracy with Adam: {accuracy(Adam_preds, Adam_labels)}")
SGD_conf_matrix = confusion_matrix(SGD_preds, SGD_labels)
Adam_conf_matrix = confusion_matrix(Adam_preds, Adam_labels)
plt.imshow(SGD_conf_matrix, cmap="hot", interpolation="nearest")
plt.title("Confusion matrix for custom SGD")
plt.show()
plt.imshow(Adam_conf_matrix, cmap="hot", interpolation="nearest")
plt.title("Confusion matrix for Adam")
plt.show()


model = BasicCNN(output_size)
print("Training CNN with custom SGD...")
custom_optimizer_sgd = SGD(model.parameters(), lr=0.1)
SGD_losses, SGD_preds, SGD_labels = train(model, custom_optimizer_sgd, batch_size, epochs)

print("Training CNN with Adam...")
optimizer_adam = Adam(model.parameters(), lr=0.01)
Adam_losses, Adam_preds, Adam_labels = train(model, optimizer_adam, batch_size=100, epochs=10)

# Plot the losses
plt.plot(SGD_losses, label="SGD")
plt.plot(Adam_losses, label="Adam")
plt.title("Loss vs Epochs for Fashion MNIST dataset")
plt.xlabel("Epoch")
plt.ylabel("Loss")
plt.legend()
plt.show()

# Plot the confusion matrix
print(f"Accuracy with custom SGD: {accuracy(SGD_preds, SGD_labels)}")
print(f"Accuracy with Adam: {accuracy(Adam_preds, Adam_labels)}")
SGD_conf_matrix = confusion_matrix(SGD_preds, SGD_labels)
Adam_conf_matrix = confusion_matrix(Adam_preds, Adam_labels)
plt.imshow(SGD_conf_matrix, cmap="hot", interpolation="nearest")
plt.title("Confusion matrix for custom SGD")
plt.show()
plt.imshow(Adam_conf_matrix, cmap="hot", interpolation="nearest")
plt.title("Confusion matrix for Adam")
plt.show()