-
Notifications
You must be signed in to change notification settings - Fork 0
/
resnet.py
72 lines (66 loc) · 2.51 KB
/
resnet.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
from __future__ import print_function
import util
import argparse
from keras.callbacks import ModelCheckpoint
parser = argparse.ArgumentParser(description='CIFAR100 Resnet script for PAP experiment')
parser.add_argument('activation', type=str, help='activation function')
parser.add_argument('-b', '--batchsize', default=128,
type=int, help='batch size')
parser.add_argument('-e', '--epochs', default=100,
type=int, help='epochs')
parser.add_argument('-l', '--lr', default=0.001,
type=float, help='learning_rate')
parser.add_argument('-s', '--seed', default=32,
type=int, help='seed for resnet')
parser.add_argument('-d', '--dataset', default='cifar100',
type=str, help='Dataset')
args = parser.parse_args()
activation = args.activation
batch_size = args.batchsize
learning_rate = args.lr
epochs = args.epochs
seed = args.seed
dataset = args.dataset.upper()
results_file = './{}-resnet-seed{}_{}-{}.csv'.format(dataset, seed, activation, learning_rate)
initialization = util.get_init_for_activation(activation)
print()
print('/==========================\\')
print("| Dataset: {} (Resnet)".format(dataset))
print("| Activation: {}".format(activation))
print("| Learning rate: {}".format(learning_rate))
print("| Batch size: {}".format(batch_size))
print("| Epochs: {}".format(epochs))
print("| Seed: {}".format(seed))
print("| Initialization: {}".format(initialization))
print('\\==========================/')
print()
if dataset == 'CIFAR100':
X_train, X_test, Y_train, Y_test = util.get_cifar100()
dims = 100
elif dataset == 'CIFAR10':
X_train, X_test, Y_train, Y_test = util.get_cifar10()
dims = 10
else:
import sys
print('Invalid dataset: {}'.format(dataset))
sys.exit(1)
print("Building...")
resnet = util.build_resnet_34(activation, initialization, dims)
print("Compiling...")
util.compile_resnet(resnet, learning_rate)
cb = util.PersistentHistory(results_file)
cb2 = ModelCheckpoint('./cifar100-resnet_seed{}_{}-{}.weights'.format(seed, activation, learning_rate), monitor='val_acc', verbose=0, save_best_only=True, mode='auto')
print("Fitting...")
resnet.fit({
'input':X_train,
'output':Y_train
},
batch_size=batch_size,
nb_epoch=epochs,
shuffle=True,
show_accuracy=True,
validation_data={
'input': X_test,
'output': Y_test
},
callbacks=[cb, cb2])