-
Notifications
You must be signed in to change notification settings - Fork 0
/
train_utils.py
183 lines (134 loc) · 6.6 KB
/
train_utils.py
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
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
import json
import torch
from torch import nn
from torchvision import datasets, models, transforms
def load_data(data_dir = './flowers/'):
train_dir = data_dir + '/train'
valid_dir = data_dir + '/valid'
test_dir = data_dir + '/test'
train_transforms = transforms.Compose([transforms.RandomRotation(30),
transforms.RandomResizedCrop(224),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize(mean=(0.485, 0.456, 0.406),
std=(0.229, 0.224, 0.225))
])
test_transforms = transforms.Compose([transforms.Resize(255),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225])
])
validation_transforms = transforms.Compose([transforms.Resize(255),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225])
])
# Load the datasets with the help of ImageFolder
train_data = datasets.ImageFolder(train_dir, transform=train_transforms)
test_data = datasets.ImageFolder(test_dir, transform=test_transforms)
validate_data = datasets.ImageFolder(valid_dir, transform=validation_transforms)
# Using the datasets above and the trainforms, define the dataloaders
train_dataloader = torch.utils.data.DataLoader(train_data, batch_size=32, shuffle=True)
test_dataloader = torch.utils.data.DataLoader(test_data, batch_size=32, shuffle=True)
validate_dataloader = torch.utils.data.DataLoader(validate_data, batch_size=32, shuffle=True)
return train_data, train_dataloader, test_dataloader, validate_dataloader
def load_json_data():
with open('cat_to_name.json', 'r') as f:
cat_to_name = json.load(f)
return cat_to_name
# setup the model, optimizer, criterion
def create_model(model_name='vgg16', hidden_input=1024,
learning_rate=0.001, mode='gpu'):
cat_to_name = load_json_data()
if model_name == 'vgg16':
model = models.vgg16(pretrained=True)
elif model_name == 'densenet121':
model = models.densenet121(pretrained=True)
elif model_name == 'alexnet':
model = models.alexnet(pretrained = True)
else:
print("Im sorry but {} is not a valid model. Did you mean vgg16, densenet121, or alexnet?".format(model_name))
for parameter in model.parameters():
parameter.requires_grad = False
#Hyper parameters
input_size = model.classifier[0].in_features
hidden_inputs = [2048, hidden_input]
output_size = len(cat_to_name)
# Build a feed-forward network
classifier = nn.Sequential(nn.Linear(input_size, hidden_inputs[0]),
nn.ReLU(),
nn.Dropout(p=0.15),
nn.Linear(hidden_inputs[0], hidden_inputs[1]),
nn.ReLU(),
nn.Dropout(p=0.15),
nn.Linear(hidden_inputs[1], output_size),
nn.LogSoftmax(dim=1))
model.classifier = classifier
if torch.cuda.is_available() and mode == 'gpu':
model.cuda()
criterion = nn.NLLLoss()
optimizer = torch.optim.Adam(model.classifier.parameters(), lr=learning_rate)
return model, optimizer, criterion
# validate the model
def validation(model, criterion, mode, validate_dataloader):
test_loss = 0
accuracy = 0
for images, labels in iter(validate_dataloader):
if torch.cuda.is_available() and mode == 'gpu':
images = images.to('cuda')
labels = labels.to('cuda')
output = model.forward(images)
test_loss += criterion(output, labels).item()
ps = torch.exp(output)
equality = (labels.data == ps.max(dim=1)[1])
accuracy += equality.type(torch.FloatTensor).mean()
return test_loss, accuracy
# trains model
def train_model(model, optimizer, criterion, train_dataloader,
validate_dataloader, epochs=5, mode='gpu'):
print("Training the model\n")
print_every = 50
steps = 0
for e in range(epochs):
running_loss = 0
for images, labels in iter(train_dataloader):
model.train()
steps += 1
if torch.cuda.is_available() and mode == 'gpu':
images = images.to('cuda')
labels = labels.to('cuda')
optimizer.zero_grad()
# Forward and backward passes
output = model.forward(images)
loss = criterion(output, labels)
loss.backward()
optimizer.step()
running_loss += loss.item()
if steps % print_every == 0:
model.eval()
with torch.no_grad():
test_loss, accuracy = validation(model, criterion, mode, validate_dataloader)
print("Epoch: {}/{}... ".format(e+1, epochs),
"Loss: {:.4f}".format(running_loss/print_every),
"Test Loss: {:.3f}...".format(test_loss/len(validate_dataloader)),
"Accuracy: {:.3f}...".format(accuracy/len(validate_dataloader))
)
running_loss = 0
model.train()
print("Finished training the model\n")
# Save the checkpoint
def save_checkpoint(model, args, optimizer, train_data):
print("Our model: \n\n", model, '\n')
print("The state dict keys: \n\n", model.state_dict().keys())
model.class_to_idx = train_data.class_to_idx
checkpoint = {'epochs': args.epochs,
'model_name': args.model_name,
'classifier': model.classifier,
'state_dict': model.state_dict(),
'optimizer_dict': optimizer.state_dict(),
'class_to_idx': model.class_to_idx,
}
torch.save(checkpoint, args.save_dir)
print("Saved the model as checkpoint\n")