运行cnn.py时报错

[LuM2016]

Traceback (most recent call last):

File "", line 1, in
debugfile('H:/DeepLearning/cnn_minst_keras/cnn.py', wdir='H:/DeepLearning/cnn_minst_keras')

File "D:\Anaconda2\lib\site-packages\spyder\utils\site\sitecustomize.py", line 888, in debugfile
debugger.run("runfile(%r, args=%r, wdir=%r)" % (filename, args, wdir))

File "D:\Anaconda2\lib\bdb.py", line 400, in run
exec cmd in globals, locals

File "", line 1, in

File "D:\Anaconda2\lib\site-packages\spyder\utils\site\sitecustomize.py", line 866, in runfile
execfile(filename, namespace)

File "D:\Anaconda2\lib\site-packages\spyder\utils\site\sitecustomize.py", line 87, in execfile
exec(compile(scripttext, filename, 'exec'), glob, loc)

File "h:/deeplearning/cnn_minst_keras/cnn.py", line 9, in
'''

File "D:\Anaconda2\lib\site-packages\keras\models.py", line 308, in add
output_tensor = layer(self.outputs[0])

File "D:\Anaconda2\lib\site-packages\keras\engine\topology.py", line 487, in call
self.build(input_shapes[0])

File "D:\Anaconda2\lib\site-packages\keras\layers\core.py", line 695, in build
name='{}_W'.format(self.name))

File "D:\Anaconda2\lib\site-packages\keras\initializations.py", line 36, in normal
return K.random_normal_variable(shape, 0.0, scale, name=name)

File "D:\Anaconda2\lib\site-packages\keras\backend\theano_backend.py", line 145, in random_normal_variable
return variable(np.random.normal(loc=0.0, scale=scale, size=shape),

File "mtrand.pyx", line 1903, in mtrand.RandomState.normal (numpy\random\mtrand\mtrand.c:18479)

File "mtrand.pyx", line 234, in mtrand.cont2_array_sc (numpy\random\mtrand\mtrand.c:3092)

ValueError: negative dimensions are not allowed

[wepe]

这份代码的backend使用的是theano，最新版keras默认后端tf。检查一下 .keras/keras.json 这份文件，参考 : #13 (comment)

[LuM2016]

感谢，楼主的回答。程序是在后端为theano的情况下运行的。

[wepe]

C盘下找.keras/keras.json，文件内容必须是：

{
    "image_dim_ordering": "th",
    "epsilon": 1e-07,
    "floatx": "float32",
    "backend": "theano"
}

[LuM2016]

十分感谢楼主的解答。问题集解决，但是程序运行到model.fit(), python.exe 停止运行。初接触keras，还请楼主指点指点。谢谢。

[oyb001]

我在centos 7.2 python 2.7环境执行，报以下错误，楼主之前使用的具体环境和模块版本是否可以提供一下。
[root@sz-xh_42f-op-test-cmdb-kafka keras]# python cnn.py
Using Theano backend.
WARNING (theano.configdefaults): g++ not detected ! Theano will be unable to execute optimized C-implementations (for both CPU and GPU) and will default to Python implementations. Performance will be severely degraded. To remove this warning, set Theano flags cxx to an empty string.
/usr/lib64/python2.7/site-packages/scipy/init.py:110: UserWarning: Numpy 1.8.2 or above is recommended for this version of scipy (detected version 1.7.1)
UserWarning)
/usr/lib64/python2.7/site-packages/numpy/lib/utils.py:254: FutureWarning: Numpy has detected that you (may be) writing to an array returned
by numpy.diagonal or by selecting multiple fields in a record
array. This code will likely break in the next numpy release --
see numpy.diagonal or arrays.indexing reference docs for details.
The quick fix is to make an explicit copy (e.g., do
arr.diagonal().copy() or arr[['f0','f1']].copy()).
ai = a.array_interface
RuntimeError: module compiled against API version 9 but this version of numpy is 7
WARNING (theano.tensor.blas): Failed to import scipy.linalg.blas, and Theano flag blas.ldflags is empty. Falling back on slower implementations for dot(matrix, vector), dot(vector, matrix) and dot(vector, vector) (numpy.core.multiarray failed to import)
RuntimeError: module compiled against API version 9 but this version of numpy is 7
RuntimeError: module compiled against API version 9 but this version of numpy is 7
Traceback (most recent call last):
File "cnn.py", line 13, in
from keras.preprocessing.image import ImageDataGenerator
File "/usr/lib64/python2.7/site-packages/keras/preprocessing/image.py", line 10, in
from scipy import linalg
File "/usr/lib64/python2.7/site-packages/scipy/linalg/init.py", line 175, in
from .misc import *
File "/usr/lib64/python2.7/site-packages/scipy/linalg/misc.py", line 5, in
from .blas import get_blas_funcs
File "/usr/lib64/python2.7/site-packages/scipy/linalg/blas.py", line 155, in
from scipy.linalg import _fblas
ImportError: cannot import name _fblas

[junlulocky]

Hey @oyb001 ,

From your description, some comments, but not necessarily useful, I guess your python cannot find the package route correctly:

for windows:
sudo /bin/env pip uninstall name _fblas ## important
sudo /bin/env python pip install name _fblas

for MAC
sudo /usr/bin/env pip uninstall name _fblas ## important
sudo /usr/bin/env python pip install name _fblas

checkout this link for why use '/usr/bin/env' beforehand: http://www.junlulocky.com/blog/pythontricks

[junlulocky]

oops, please remove 'name' before _fblas.

[oyb001]

@junlulocky ,python里面没有_fblas这边包
[root@sz-xh_42f-op-test-cmdb-kafka ~]# pip install blas
Collecting blas
Could not find a version that satisfies the requirement blas (from versions: )
No matching distribution found for blas
[root@sz-xh_42f-op-test-cmdb-kafka ~]# pip freeze |grep _fblas
[root@sz-xh_42f-op-test-cmdb-kafka ~]# pip install _fblas
Invalid requirement: '_fblas'
Traceback (most recent call last):
File "/usr/lib/python2.7/site-packages/pip-9.0.1-py2.7.egg/pip/req/req_install.py", line 82, in init
req = Requirement(req)
File "/usr/lib/python2.7/site-packages/pip-9.0.1-py2.7.egg/pip/_vendor/packaging/requirements.py", line 96, in init
requirement_string[e.loc:e.loc + 8]))
InvalidRequirement: Invalid requirement, parse error at "'_fblas'"

[root@sz-xh_42f-op-test-cmdb-kafka ~]#

[oyb001]

@junlulocky ,问题已解决，
步骤1：卸载Keras Theano Tensorflow
步骤2: 升级numpy至numpy-1.13.3版本

本人安装环境:
os:centos 7.2
python:2.7.5
theano: 0.8.2
tensorflow: 1.0.1
Using TensorFlow backend.
keras: 1.2.2
numpy:1.13.3
scipy==0.19.1