-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtraining_utils.py
executable file
·181 lines (140 loc) · 5.43 KB
/
training_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
import torch
from tqdm import tqdm
# from tqdm.notebook import tqdm_notebook
import torch.nn.functional as F
import numpy as np
from sklearn.metrics import accuracy_score, classification_report
# def per_class_accuracy(labels, predictions):
# """
# Calculates the per-class accuracy given two arrays of labels and predictions
# """
#
# labels = labels.tolist()
# predictions = predictions.tolist()
#
# class_report = classification_report(labels, predictions, output_dict=True, zero_division=1)
# accuracy_per_class = {}
# for class_name in class_report.keys():
# if class_name.isdigit():
# class_id = int(class_name)
# # accuracy_per_class[class_id] = class_report[class_name]['recall']
# accuracy_per_class[class_id] = class_report[class_name]['precision']
# # print(class_report[class_name])
# return np.array(list(accuracy_per_class.values()), dtype='float64')
def get_per_cls_imgs(labels, num_cls):
labels = labels.tolist()
per_cls_imgs = np.zeros(num_cls)
for i in range(num_cls):
per_cls_imgs[i] = labels.count(int(i))
# print(f'Labels - {labels}')
# print(f'per_cls_imgs - {per_cls_imgs}')
return np.array(per_cls_imgs)
def get_per_cls_correct(labels, predictions, num_cls):
"""
Calculates number of images correctly classified per class count
"""
labels = labels.tolist()
predictions = predictions.tolist()
per_cls_correct = np.zeros(num_cls)
batch_size = len(labels)
for i in range(batch_size):
class_id = labels[i]
if labels[i] == predictions[i]:
per_cls_correct[class_id] += 1
else:
pass
# print(f'Predictions - {predictions}')
# print(f'per_cls_correct - {per_cls_correct}')
return np.array(per_cls_correct)
# def get_per_cls_acc(preds, labels, num_cls):
#
# print(f'Preds - {preds.tolist()}')
# print(f'Labels - {labels.tolist()}')
#
# preds = np.array(preds)
# labels = np.array(labels)
#
# num_images = len(labels)
#
# assert len(labels) == len(preds), 'Number of labels and predictions not matching !'
#
# per_cls_acc = np.zeros(num_cls)
# for i in range(num_cls):
# per_cls_acc[i] =
#
# return per_cls_acc
# Training function.
def train(model, trainloader, optimizer, criterion, device, num_cls):
model.train()
print('Training')
train_running_loss = 0.0
train_running_correct = 0
per_cls_correct = np.zeros(num_cls, dtype='float64')
per_cls_imgs = np.zeros(num_cls, dtype='float64')
counter = 0
for i, data in tqdm(enumerate(trainloader), total=len(trainloader)):
counter += 1
image, labels = data
image = image.to(device)
labels = labels.to(device)
optimizer.zero_grad()
# Forward pass.
outputs = model(image)
# # Added softmax layer before computing loss
# outputs = F.softmax(outputs, dim=1)
# Calculate the loss.
loss = criterion(outputs, labels)
train_running_loss += loss.item()
# Calculate the accuracy.
_, preds = torch.max(outputs.data, 1)
train_running_correct += (preds == labels).sum().item()
# Backpropagation
loss.backward()
# Update the weights.
optimizer.step()
per_cls_correct += get_per_cls_correct(labels, preds, num_cls)
per_cls_imgs += get_per_cls_imgs(labels, num_cls)
# print(f'per_cls_correct - {per_cls_correct}')
# print(f'per_cls_imgs - {per_cls_imgs}')
# Loss and accuracy for the complete epoch.
epoch_loss = train_running_loss / counter
# epoch_acc = 100. * (train_running_correct / len(trainloader.dataset))
epoch_acc = 100. * (train_running_correct / len(trainloader.dataset))
per_cls_acc = per_cls_correct * 100 / per_cls_imgs
return epoch_loss, epoch_acc, model, per_cls_acc
# Validation function.
def validate(model, testloader, criterion, device, num_cls):
model.eval()
print('Validation')
valid_running_loss = 0.0
valid_running_correct = 0
per_cls_correct = np.zeros(num_cls, dtype='float64')
per_cls_imgs = np.zeros(num_cls, dtype='float64')
counter = 0
with torch.no_grad():
for i, data in tqdm(enumerate(testloader), total=len(testloader)):
counter += 1
image, labels = data
# print(f'Image shape - {image.shape}')
# print(f'Image type - {type(image)}')
# print(f'Labels - {labels}')
# print(f'Labels shape - {labels.shape}')
image = image.to(device)
labels = labels.to(device)
# Forward pass.
outputs = model(image)
# Calculate the loss.
loss = criterion(outputs, labels)
valid_running_loss += loss.item()
# Calculate the accuracy.
_, preds = torch.max(outputs.data, 1)
valid_running_correct += (preds == labels).sum().item()
per_cls_correct += get_per_cls_correct(labels, preds, num_cls)
per_cls_imgs += get_per_cls_imgs(labels, num_cls)
# print(f'per_cls_correct - {per_cls_correct}')
# print(f'per_cls_imgs - {per_cls_imgs}')
# Loss and accuracy for the complete epoch.
epoch_loss = valid_running_loss / counter
epoch_acc = 100. * (valid_running_correct / len(testloader.dataset))
per_cls_acc = per_cls_correct * 100 / per_cls_imgs
return epoch_loss, epoch_acc, model, per_cls_acc