forked from jacobandreas/nmn
-
Notifications
You must be signed in to change notification settings - Fork 0
/
bug.py
117 lines (103 loc) · 3.57 KB
/
bug.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
#!/usr/bin/env python2
import caffe
import apollocaffe
from apollocaffe import ApolloNet, layers
import numpy as np
import timeit
#caffe.set_mode_gpu()
apollocaffe.set_device(0)
net = ApolloNet()
batch_size = 64
data = np.random.random(size=(batch_size, 512, 20, 20)).astype(np.float32)
labels = np.random.randint(10, size=(batch_size,)).astype(np.int32).astype(np.float32)
#print data.dtype
#print labels.dtype
#def load_mem():
# net.clear_forward()
# net.f(layers.MemoryData(
# "mem", data, labels, tops=["input_top", "label_top"],
# batch_size=batch_size, channels=512, width=20, height=20))
#
#def load_np():
# net.clear_forward()
# net.f(layers.NumpyData("np", data))
#
#load_mem()
#load_np()
#data = np.zeros((64, 512, 20, 20))
for i in range(10):
print
net.clear_forward()
import time; s = time.time()
net.f(layers.NumpyData('a', data))
print time.time() - s
import time; s = time.time()
net.clear_forward()
net.blobs['a'].data[:] = data
print time.time() - s
import time; s = time.time()
net.clear_forward()
net.f(layers.MemoryData(
"b", data, labels, tops=["input_top", "label_top"],
batch_size=batch_size, channels=512, width=20, height=20))
print time.time() - s
#def prep():
# net.clear_forward()
# net.f(layers.NumpyData("input", data=np.random.random(size=(batch_size,512,20,20))))
#
#def load_layer():
# net.clear_forward()
# net.blobs["input"].data[...] = np.random.random(size=(batch_size,512,20,20))
# net.f(layers.InnerProduct("ip", 512, bottoms=["input"]))
#
#def load_sloppy():
# net.clear_forward()
# net.f(layers.NumpyData("input", data=np.random.random(size=(batch_size,512,20,20))))
# net.f(layers.InnerProduct("ip", 512, bottoms=["input"]))
#
#prep()
#
#print "mem", timeit.timeit("load_mem()", number=100, setup="from __main__ import load_mem")
#print "np", timeit.timeit("load_np()", number=100, setup="from __main__ import load_np")
#for i in range(32, 64):
# load(i)
# time = timeit.timeit('load(%d)' % i, number=100, setup="from __main__ import load")
# print time / i
#net.clear_forward()
#net.f(layers.NumpyData("words", data=np.random.randint(10, size=(7,))))
#print net.blobs["words"].shape
#net.f(layers.Wordvec("vecs", 10, 10, bottoms=["words"]))
#print net.blobs["vecs"].shape
#
#net.clear_forward()
#net.f(layers.NumpyData("words", data=np.random.randint(10, size=(14,))))
#print net.blobs["words"].shape
#net.f(layers.Wordvec("vecs", 10, 10, bottoms=["words"]))
#print net.blobs["vecs"].shape
#@profile
#def main():
# for i in range(1000):
# net.clear_forward()
# net.f(layers.NumpyData("input", data=np.random.random((256, 8, 20, 20))))
# net.f(layers.NumpyData("target", data=np.random.randint(32, size=(256,))))
# net.f(layers.InnerProduct("ip1", 32, bottoms=["input"]))
# net.f(layers.SoftmaxWithLoss("loss", bottoms=["ip1", "target"]))
# net.backward()
#
# net.clear_forward()
# net.f(layers.NumpyData("input", data=np.random.random((256, 8, 20, 20))))
# net.f(layers.NumpyData("target", data=np.random.randint(32, size=(256,))))
# net.f(layers.InnerProduct("ip2", 32, bottoms=["input"]))
# net.f(layers.SoftmaxWithLoss("loss2", bottoms=["ip2", "target"]))
# net.backward()
#
#main()
#print net.blobs["output"].data
#
#net.clear_forward()
#net.f(layers.NumpyData("input", data=np.ones((1, 16, 2, 2))))
#net.f(layers.Convolution("conv2", (1,1), 1, bottoms=["input"]))
#net.f(layers.InnerProduct("output", 8, bottoms=["conv2"]))
#net.backward()
#
#print net.blobs["output"].data