-
Notifications
You must be signed in to change notification settings - Fork 2
/
train.py
191 lines (146 loc) · 6.9 KB
/
train.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
184
185
186
187
188
189
190
191
import torch
import torch.nn as nn
import torch.optim as optim
from torch.optim import lr_scheduler
from datasets.IemocapDataset import IemocapDataset
from models.VGG_convnet import VGG_convnet
from models.DNN import DNN
import time
import copy
from scipy import stats
def compute_number_of_corrects(preds, data, n_frames):
# Compute number of corrects for variable batch size
n_corrects = 0
for i in range(len(n_frames)):
start_idx = 0 if i == 0 else torch.sum(n_frames[:i]).long()
end_idx = n_frames[i].long()
target_label = data[start_idx]
predicted_label = stats.mode(preds[start_idx:start_idx+end_idx]).mode[0]
n_corrects += (target_label == predicted_label)
return n_corrects
def train_model_vgg(model, criterion, optimizer, scheduler, num_epochs=25):
since = time.time()
best_model_wts = copy.deepcopy(model.state_dict())
best_acc = 0.0
for epoch in range(num_epochs):
print('Epoch {}/{}'.format(epoch, num_epochs - 1))
print('-' * 10)
# Each epoch has a training and validation phase
for phase in ['train', 'val']:
if phase == 'train':
model.train() # Set model to training mode
else:
model.eval() # Set model to evaluate mode
running_loss = 0.0
running_corrects = 0
# Iterate over data.
for inputs, labels in dataloaders[phase]:
inputs = inputs.to(device)
labels = labels.long().to(device)
# zero the parameter gradients
optimizer.zero_grad()
# forward
# track history if only in train
with torch.set_grad_enabled(phase == 'train'):
outputs = model(inputs)
_, preds = torch.max(outputs, 1)
loss = criterion(outputs, labels)
# backward + optimize only if in training phase
if phase == 'train':
loss.backward()
optimizer.step()
# statistics
running_loss += loss.item() * inputs.size(0)
running_corrects += torch.sum(preds == labels.data)
if phase == 'train':
scheduler.step()
epoch_loss = running_loss / dataset_sizes[phase]
epoch_acc = running_corrects.double() / dataset_sizes[phase]
print('{} Loss: {:.4f} Acc: {:.4f}'.format(phase, epoch_loss, epoch_acc))
# deep copy the model
if phase == 'val' and epoch_acc > best_acc:
best_acc = epoch_acc
best_model_wts = copy.deepcopy(model.state_dict())
print()
time_elapsed = time.time() - since
print('Training complete in {:.0f}m {:.0f}s'.format(time_elapsed // 60, time_elapsed % 60))
print('Best val Acc: {:4f}'.format(best_acc))
# load best model weights
model.load_state_dict(best_model_wts)
return model
def train_model(model, criterion, optimizer, scheduler, num_epochs=25):
since = time.time()
best_model_wts = copy.deepcopy(model.state_dict())
best_acc = 0.0
for epoch in range(num_epochs):
print('Epoch {}/{}'.format(epoch, num_epochs - 1))
print('-' * 10)
# Each epoch has a training and validation phase
for phase in ['train', 'val']:
if phase == 'train':
model.train() # Set model to training mode
else:
model.eval() # Set model to evaluate mode
running_loss = 0.0
running_corrects = 0
total_n_frames = 0
# Iterate over data.
for inputs, emotions, n_frames in dataloaders[phase]:
inputs = inputs.to(device)
emotions = emotions.long().to(device)
# zero the parameter gradients
optimizer.zero_grad()
# forward
# track history if only in train
with torch.set_grad_enabled(phase == 'train'):
outputs = model(inputs)
_, preds = torch.max(outputs, 1)
loss = criterion(outputs, emotions)
# backward + optimize only if in training phase
if phase == 'train':
loss.backward()
optimizer.step()
# statistics
running_loss += loss.item() * inputs.size(0)
# running_corrects += torch.sum(preds == emotions.data)
running_corrects += compute_number_of_corrects(preds, emotions.data, n_frames)
total_n_frames += inputs.size(0)
if phase == 'train':
scheduler.step()
epoch_loss = running_loss / total_n_frames
epoch_acc = running_corrects.double() / dataset_sizes[phase]
print('{} Loss: {:.4f} Acc: {:.4f}'.format(phase, epoch_loss, epoch_acc))
# deep copy the model
if phase == 'val' and epoch_acc > best_acc:
best_acc = epoch_acc
best_model_wts = copy.deepcopy(model.state_dict())
print()
time_elapsed = time.time() - since
print('Training complete in {:.0f}m {:.0f}s'.format(time_elapsed // 60, time_elapsed % 60))
print('Best val Acc: {:4f}'.format(best_acc))
# load best model weights
model.load_state_dict(best_model_wts)
return model
# train on the GPU or on the CPU, if a GPU is not available
device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
# load train and valid datasets
datasets = {
'train': IemocapDataset('/home/alanwuha/Documents/Projects/datasets/iemocap/IEMOCAP_full_release', sessions=[1, 2, 4, 5]),
'val': IemocapDataset('/home/alanwuha/Documents/Projects/datasets/iemocap/IEMOCAP_full_release', sessions=[3])
}
dataset_sizes = { x: len(datasets[x]) for x in ['train', 'val'] }
dataloaders = { x: torch.utils.data.DataLoader(datasets[x], batch_size=2, shuffle=True, num_workers=4, collate_fn=IemocapDataset.collate_fn) for x in ['train', 'val'] }
# dataloaders = { x: torch.utils.data.DataLoader(datasets[x], batch_size=128, shuffle=(x=='train'), num_workers=4, collate_fn=IemocapDataset.collage_fn_vgg) for x in ['train', 'val'] }
# Model
model_ft = DNN(400, 1000, 1500, 9)
# model_ft = VGG_convnet()
model_ft = model_ft.to(device)
criterion = nn.CrossEntropyLoss()
# Observe that all parameters are being optimized
# optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)
optimizer_ft = optim.Adam(model_ft.parameters(), lr=0.001)
# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)
# Train and evaluate
model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler, num_epochs=25)
# model_ft = train_model_vgg(model_ft, criterion, optimizer_ft, exp_lr_scheduler, num_epochs=25)