no module name activation

[assadRasheed]

ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-9-86b7106a2640> in <module>()
     14 from lasagne.layers import SliceLayer as SL
     15 
---> 16 import voxnet
     17 import CAcheckpoints
     18 import GANcheckpoints

/usr/local/lib/python3.6/dist-packages/voxnet/__init__.py in <module>()
      1 from .version import __version__
      2 
----> 3 import activations
      4 import checkpoints
      5 import init

ModuleNotFoundError: No module named 'activations

[Hailey-Liu]

Same error

[tranvnhan]

Same with me.

My setup:

• Ubuntu 18.04
• torch 1.2.0

[claell]

I also get this error (No module named 'activations') on Windows 7 with Python 3 when running the Python script to convert the training data. Might this be related to the usage of Python 3?

[weixiewen]

from . import XXX

[claell]

@weixiewen Can you elaborate what you mean with this?

[weixiewen]

@weixiewen Can you elaborate what you mean with this?

under voxnet———init.py
the import error
can be wrote
from . import XXX and then fixed
XXX can be act… .py layers.py
like
from . import layers
But … yesterday and yesteryesterdays i try to change the project voxnet py2 to py3 and then trained the model it broken... so sad.
Such as cStringio to io.bytersio....
No use...Nearly read through the whole conda site-pkg _ sources of tf_core and Theano(the new) and voxnet
Cry..