-
Notifications
You must be signed in to change notification settings - Fork 0
/
train.py
149 lines (133 loc) · 5.06 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
import torch
import time, datetime
import models, utils
from utils import print
from torch.optim.lr_scheduler import StepLR
import datetime
defaultProgressViewLoop = 1000
import numpy as np
def train_model(net, data_loader, epochs, lr, device, progressViewLoop=defaultProgressViewLoop):
startTime = time.time()
net.train()
net = net.to(device)
criterion = torch.nn.MSELoss()
optimizer = torch.optim.Adam(net.parameters(), lr = lr)
# scheduler = StepLR(optimizer, step_size=10, gamma=0.5)
# optimizer = torch.optim.SGD(net.parameters(), lr=lr)
for epoch in range(epochs):
running_loss = 0.0
progressed_items = 0
for idx, batch in enumerate(data_loader):
batch = batch.to(device)
optimizer.zero_grad()
outputs = net(batch)
outputs= outputs.view(-1)
loss = criterion(outputs, batch.y)
loss.backward()
optimizer.step() # update module
running_loss += loss.item()
progressed_items += batch.num_graphs
if idx % progressViewLoop == progressViewLoop-1:
print("\n[epoch:%d, batch:%5d] loss: %.8f (time elapsed: %3f)" %
(epoch + 1, idx + 1, running_loss / progressed_items,
time.time() - startTime))
running_loss /= len(data_loader.dataset)
print("\n[epoch:%d] - finished, loss: %.8f (time elapsed: %3f time:%s)" %
(epoch + 1, running_loss, time.time() - startTime, datetime.datetime.now()))
# scheduler.step()
return net
@torch.no_grad()
def test(net, data_loader, device, status="Test", progressViewLoop=defaultProgressViewLoop):
net.eval()
net = net.to(device)
eval_loss = 0
progressed_items = 0
for idx, batch in enumerate(data_loader):
batch = batch.to(device)
outputs = net(batch)
predicted = outputs.view(-1)
target = batch.y
eval_loss += mse_loss(predicted, target).item()
progressed_items += batch.num_graphs
if idx % progressViewLoop == progressViewLoop - 1:
print("\n[Test: %d] - loss: %8f" %
(idx, float(eval_loss/progressed_items)))
acc = eval_loss / len(data_loader.dataset)
print(status + " accuracy: ", acc)
@torch.no_grad()
def test_mape(net, data_loader, device, status="Test", progressViewLoop=defaultProgressViewLoop):
net.eval()
net = net.to(device)
eval_loss = 0
progressed_items = 0
for idx, batch in enumerate(data_loader):
batch = batch.to(device)
outputs = net(batch)
predicted = outputs.view(-1)
target = batch.y
eval_loss += mape_loss(predicted, target, device).item()
progressed_items += batch.num_graphs
if idx % progressViewLoop == progressViewLoop - 1:
print("\n[Test: %d] - loss: %8f" %
(idx, float(eval_loss/progressed_items)))
acc = eval_loss / len(data_loader.dataset)
print(status + " accuracy: ", acc)
@torch.no_grad()
def testDraw(net, data_loader, device, status="Test", progressViewLoop=defaultProgressViewLoop):
from numpy import absolute
net.eval()
net = net.to(device)
for idx, batch in enumerate(data_loader):
batch = batch.to(device)
outputs = net(batch)
predicted = outputs.view(-1)
input = batch.x.cpu().detach().numpy()
labels = batch.y.cpu().detach().numpy()
nodes = batch.nds[0]
predicted = predicted.cpu().detach().numpy()
from funcs import plot
plot.drawMesh(input, nodes, labels)
plot.drawMesh(input, nodes, predicted)
plot.drawMesh(input, nodes, absolute(labels-predicted))
target = batch.y
eval_loss = mse_loss(predicted, target).item()
if idx % progressViewLoop == progressViewLoop - 1:
print("\n[Test: %d] - loss: %8f" %
(idx, float(eval_loss)))
@torch.no_grad()
def testDrawGen(net, path, device, label=None, pred="vms"):
from torch_geometric.data import DataLoader
from numpy import absolute
testList = [path]
testdata = utils.testDataset(testList, pred=pred)
testloader = DataLoader(testdata, batch_size=1, shuffle=True, num_workers=16)
net.eval()
net = net.to(device)
for idx, batch in enumerate(testloader):
batch = batch.to(device)
outputs = net(batch)
predicted = outputs.view(-1)
target = batch.y
input = batch.x.cpu().detach().numpy()
labels = batch.y.cpu().detach().numpy()
nodes = batch.nds[0]
pred = predicted.cpu().detach().numpy()
from funcs import plot
maxValue = float(np.max(labels))
plot.drawMesh(input, nodes, labels, maxValue, "Ground truth", label, True)
plot.drawMesh(input, nodes, pred, maxValue, "Prediction", label, True)
plot.drawMesh(input, nodes, absolute(labels-pred), None,"Error", label, True)
eval_loss = mse_loss(predicted, target).item()
print("\n[Test: %d] - acc: %8f" %
(idx, float(eval_loss)))
def weighted_mse_loss(input, target, weight):
return (weight * (input - target) ** 2).mean()
def mse_loss(input, target):
return ((input - target) ** 2).mean()
def mape_loss(input, target, device, min = 0.01):
epsilon = min * torch.ones(target.shape[0]).to(device)
return (torch.abs(input - target) / (target + epsilon)).mean() * 100.0
def rel_mape_loss(input, target):
stack = torch.stack([input, target], dim=0)
max = torch.max(stack, dim=0)[0]
return (torch.abs(input - target) / max).mean() * 100.0