Open zyui opened 7 years ago
Related to #234. The only workaround I found for now is to force pip to downgrade Theano to 0.7.
This issue is happening because the latest released version of Lasagne (0.1) still uses downsample from Theano, but the latest released of Theano no longer supports downsample. @zyui @a01200356 If you prefer to use the newest version of Theano, you could alternatively upgrade Lasagne to the development version via pip install --upgrade https://github.com/Lasagne/Lasagne/archive/master.zip
(a la http://lasagne.readthedocs.io/en/latest/user/installation.html)
change from theano.tensor.signal.pool import max_pool_2d to from theano.tensor.signal.pool import pool_2d
@dibidave Worked for me, thanks!
@dibidave It works for me! Thank you!
@dibidave worked for me, thank you!
@dibidave ..thanks its work for me..
@dibidave It works for me, thank you!
@dibidave, Thanks..that worked
Worked for me too, thanks, @dibidave.
@dibidave Thanks! It worked
Thank you so much @dibidave It worked .
Sorry to reask this question again:
I installed the theano with version of 1.0.2, but I can not import the downsample module, I saw several posts for the same issue, and I tried to upgrade the version of Lasagne pip install --upgrade https://github.com/Lasagne/Lasagne/archive/master.zip
, but it did not work for me.
I saw that the question was asked with theano for version of 0.8. So I wonder if anyone has the same issue for the lasted version of theano and has some workaround?
Thanks in advance
pip install -I theano==0.7
for anyone who needs it
@dibidave Thanks a lot!!!
@dibidave that works for me too. Thanks a lot!
You just need to install Theano & Lasagne again by following commands.
pip install --upgrade https://github.com/Theano/Theano/archive/master.zip
pip install --upgrade https://github.com/Lasagne/Lasagne/archive/master.zip
`from sknn import ae import numpy as np import warnings warnings.filterwarnings("ignore", category=DeprecationWarning)
x = np.asarray(range(0,16))
nn = ae.AutoEncoder( layers=[ ae.Layer("Sigmoid", units=4), ae.Layer("Sigmoid", units=8), ae.Layer("Sigmoid", units=4) ], learning_rate=0.02) nn.fit(x)`
And then comes the arror_Traceback (most recent call last): File "/Users/zzzyui/Desktop/CSE514A/hw4/ae1.py", line 17, in
learning_rate=0.02)
File "/Library/Python/2.7/site-packages/sknn/nn.py", line 513, in init
self._setup()
File "/Library/Python/2.7/site-packages/sknn/ae.py", line 100, in setup
backend.setup()
File "/Library/Python/2.7/site-packages/sknn/backend/init.py", line 24, in setup
from . import lasagne
File "/Library/Python/2.7/site-packages/sknn/backend/lasagne/init.py", line 5, in
from .mlp import MultiLayerPerceptronBackend
File "/Library/Python/2.7/site-packages/sknn/backend/lasagne/mlp.py", line 25, in
import lasagne.layers
File "/Library/Python/2.7/site-packages/lasagne/init.py", line 19, in
from . import layers
File "/Library/Python/2.7/site-packages/lasagne/layers/init.py", line 7, in
from .pool import *
File "/Library/Python/2.7/site-packages/lasagne/layers/pool.py", line 6, in
from theano.tensor.signal import downsample
ImportError: cannot import name downsample