-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata.py
142 lines (133 loc) · 4.83 KB
/
data.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
import os
import torch
import torchvision.transforms as transforms
import torchvision.datasets as datasets
import torch.utils.data as data_utils
num_classes_dict = {
"CIFAR10":10,
"CIFAR100":100,
"DOGFISH":1,
}
def get_data(dataset, data_path, batch_size, num_workers):
assert dataset in ["CIFAR10", "CIFAR100","DOGFISH"]
print('Loading dataset {} from {}'.format(dataset, data_path))
if dataset in ["CIFAR10", "CIFAR100"]:
ds = getattr(datasets, dataset.upper())
path = os.path.join(data_path, dataset.lower())
transform_train = transforms.Compose([
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
transform_test = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
train_set = ds(path, train=True, download=True, transform=transform_train)
#print("here is get_data")
#print(train_set)
val_set = ds(path, train=True, download=True, transform=transform_test)
test_set = ds(path, train=False, download=True, transform=transform_test)
train_sampler = None
val_sampler = None
elif dataset in ["DOGFISH"]:
path = os.path.join(data_path, dataset.lower())
import numpy as np
FilePath="./features/dog_fish_features.npz"
print('Loading dogfish features from disk...')
f=np.load(FilePath)
X_train = f['X_train']
X_test = f['X_test']
Y_train = f['Y_train']
print(Y_train)
Y_test = f['Y_test']
#print(X_train.shape)
X_train = torch.Tensor(list(X_train))
Y_train = torch.Tensor(list(Y_train))
X_test = torch.Tensor(list(X_test))
Y_test = torch.Tensor(list(Y_test))
train_set = data_utils.TensorDataset(X_train, Y_train)
test_set = data_utils.TensorDataset(X_test, Y_test)
val_set = data_utils.TensorDataset(X_train, Y_train)
train_sampler = None
val_sampler = None
else:
raise Exception("Invalid dataset %s"%dataset)
loaders = {
'train': torch.utils.data.DataLoader(
train_set,
batch_size=batch_size,
shuffle=(train_sampler is None),
sampler=train_sampler,
num_workers=num_workers,
pin_memory=True
),
'val': torch.utils.data.DataLoader(
train_set,
batch_size=batch_size,
sampler=val_sampler,
num_workers=num_workers,
pin_memory=True
),
'test': torch.utils.data.DataLoader(
test_set,
batch_size=batch_size,
shuffle=False,
num_workers=num_workers,
pin_memory=True
)
}
return loaders
'''
def get_data(dataset, data_path, batch_size, num_workers):
assert dataset in ["CIFAR10", "CIFAR100"]
print('Loading dataset {} from {}'.format(dataset, data_path))
if dataset in ["CIFAR10", "CIFAR100"]:
ds = getattr(datasets, dataset.upper())
path = os.path.join(data_path, dataset.lower())
transform_train = transforms.Compose([
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
transform_test = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])
train_set = ds(path, train=True, download=True, transform=transform_train)
print("here is get_data")
print(train_set)
val_set = ds(path, train=True, download=True, transform=transform_test)
test_set = ds(path, train=False, download=True, transform=transform_test)
train_sampler = None
val_sampler = None
else:
raise Exception("Invalid dataset %s"%dataset)
loaders = {
'train': torch.utils.data.DataLoader(
train_set,
batch_size=batch_size,
shuffle=(train_sampler is None),
sampler=train_sampler,
num_workers=num_workers,
pin_memory=True
),
'val': torch.utils.data.DataLoader(
train_set,
batch_size=batch_size,
sampler=val_sampler,
num_workers=num_workers,
pin_memory=True
),
'test': torch.utils.data.DataLoader(
test_set,
batch_size=batch_size,
shuffle=False,
num_workers=num_workers,
pin_memory=True
)
}
return loaders
'''