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
|
#
# Copyright (C) 2017 The Android Open Source Project
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# model
model = Model()
i1 = Input("op1", "TENSOR_FLOAT32", "{1, 2, 3, 2}") # input tensor 0
i2 = Input("op2", "TENSOR_FLOAT32", "{1, 2, 3, 2}") # input tensor 1
i3 = Input("op3", "TENSOR_FLOAT32", "{1, 2, 3, 2}") # input tensor 2
axis0 = Int32Scalar("axis0", 3)
r = Output("result", "TENSOR_FLOAT32", "{1, 2, 3, 6}") # output
model = model.Operation("CONCATENATION", i1, i2, i3, axis0).To(r)
# Example 1.
input0 = {i1: [-0.03203143, -0.0334147 , -0.02527265, 0.04576106, 0.08869292,
0.06428383, -0.06473722, -0.21933985, -0.05541003, -0.24157837,
-0.16328812, -0.04581105],
i2: [-0.0569439 , -0.15872048, 0.02965238, -0.12761882, -0.00185435,
-0.03297619, 0.03581043, -0.12603407, 0.05999133, 0.00290503,
0.1727029 , 0.03342071],
i3: [ 0.10992613, 0.09185287, 0.16433905, -0.00059073, -0.01480746,
0.0135175 , 0.07129054, -0.15095694, -0.04579685, -0.13260484,
-0.10045543, 0.0647094 ]}
output0 = {r: [-0.03203143, -0.0334147 , -0.0569439 , -0.15872048, 0.10992613,
0.09185287, -0.02527265, 0.04576106, 0.02965238, -0.12761882,
0.16433905, -0.00059073, 0.08869292, 0.06428383, -0.00185435,
-0.03297619, -0.01480746, 0.0135175 , -0.06473722, -0.21933985,
0.03581043, -0.12603407, 0.07129054, -0.15095694, -0.05541003,
-0.24157837, 0.05999133, 0.00290503, -0.04579685, -0.13260484,
-0.16328812, -0.04581105, 0.1727029 , 0.03342071, -0.10045543,
0.0647094 ]}
# Instantiate an example
Example((input0, output0))
'''
# The above data was generated with the code below:
with tf.Session() as sess:
t1 = tf.random_normal([1, 2, 3, 2], stddev=0.1, dtype=tf.float32)
t2 = tf.random_normal([1, 2, 3, 2], stddev=0.1, dtype=tf.float32)
t3 = tf.random_normal([1, 2, 3, 2], stddev=0.1, dtype=tf.float32)
c1 = tf.concat([t1, t2, t3], axis=3)
print(c1) # print shape
print( sess.run([tf.reshape(t1, [12]),
tf.reshape(t2, [12]),
tf.reshape(t3, [12]),
tf.reshape(c1, [1*2*3*(2*3)])]))
'''
|
