質問編集履歴
1
上方修正
title
CHANGED
|
File without changes
|
body
CHANGED
|
@@ -1,49 +1,1 @@
|
|
|
1
|
-
AssertionError: Incoming Tensor shape must be 3-D とエラーが出ました。
|
|
1
|
+
AssertionError: Incoming Tensor shape must be 3-D とエラーが出ました。
|
|
2
|
-
Tracebackには
|
|
3
|
-
```ここに言語を入力
|
|
4
|
-
Traceback (most recent call last):
|
|
5
|
-
File "cnn.py", line 154, in <module>
|
|
6
|
-
net = conv_1d(net, 4, 16, activation='relu')
|
|
7
|
-
File "/Users/xxx/anaconda/xxx/lib/python2.7/site-packages/tflearn/layers/conv.py", line 802, in conv_1d
|
|
8
|
-
assert len(input_shape) == 3, "Incoming Tensor shape must be 3-D"
|
|
9
|
-
AssertionError: Incoming Tensor shape must be 3-D
|
|
10
|
-
```
|
|
11
|
-
とエラーが出ました。
|
|
12
|
-
コードには、
|
|
13
|
-
```ここに言語を入力
|
|
14
|
-
# coding: utf-8
|
|
15
|
-
import tensorflow as tf
|
|
16
|
-
import tflearn
|
|
17
|
-
|
|
18
|
-
from tflearn.layers.core import input_data,dropout,fully_connected
|
|
19
|
-
from tflearn.layers.conv import conv_3d, max_pool_3d
|
|
20
|
-
from tflearn.layers.normalization import local_response_normalization
|
|
21
|
-
from tflearn.layers.estimator import regression
|
|
22
|
-
|
|
23
|
-
tf.reset_default_graph()
|
|
24
|
-
net = input_data(shape=[20000, 4, 42])
|
|
25
|
-
net = conv_1d(net, 4, 16, activation='relu')
|
|
26
|
-
net = max_pool_1d(net, 1)
|
|
27
|
-
net = tflearn.activations.relu(net)
|
|
28
|
-
net = dropout(net, 0.5)
|
|
29
|
-
net = tflearn.fully_connected(net, 2, activation='softmax')
|
|
30
|
-
net = tflearn.regression(net, optimizer='adam', learning_rate=0.5, loss='categorical_crossentropy')
|
|
31
|
-
|
|
32
|
-
model = tflearn.DNN(net)
|
|
33
|
-
|
|
34
|
-
```
|
|
35
|
-
と書きました。
|
|
36
|
-
```ここに言語を入力
|
|
37
|
-
net = conv_3d(net, 4, 16, activation='relu')
|
|
38
|
-
net = max_pool_3d(net, 1)
|
|
39
|
-
```
|
|
40
|
-
と書き換えると、
|
|
41
|
-
```ここに言語を入力
|
|
42
|
-
Traceback (most recent call last):
|
|
43
|
-
File "cnn.py", line 154, in <module>
|
|
44
|
-
net = conv_3d(net, 4, 16, activation='relu')
|
|
45
|
-
File "/Users/xxx/anaconda/xxx/lib/python2.7/site-packages/tflearn/layers/conv.py", line 1011, in conv_3d
|
|
46
|
-
assert len(input_shape) == 5, "Incoming Tensor shape must be 5-D"
|
|
47
|
-
AssertionError: Incoming Tensor shape must be 5-D
|
|
48
|
-
```
|
|
49
|
-
のようにエラーが出ました。何が問題でどう修正したら良いのでしょうか？
|