Closed tobimichigan closed 3 years ago
JingqingZ
commented
5 years ago Hi, it seems tensorlayer is not in your conda environment.
To add a path to the PATH, you need to find ~/.bashrc and add the path /home/hgh/.local/bin into it.
Please check https://stackoverflow.com/questions/14637979/how-to-permanently-set-path-on-linux-unix
se7enXF
commented
5 years ago Obviously, it is caused by different python environment. If you use conda make sure that you are always in conda environment. Check your command line if you are using correct environment when you use pip or python.
tobimichigan
commented
5 years ago @JingqingZ & @se7enXF , for the record, I use the following on my system running, flawlessly: keras gpu [all the modules for scientific computation] tensorflow_gpu[with all the features for computation] pytorch_gpu[with all the features for computation]
My question is why should Tensorlayer be finding it difficult to be found in an environment that other python modules find so easily to run?
se7enXF
commented
5 years ago Because you install tensorlayer in python3.7, but install keras and tensorflow in conda python3.6. From your upper information:
# In your system python3.7
Requirement already satisfied: tensorlayer in ./.local/lib/python3.7/site-packages (2.1.0)
# In your conda python3.6
conda list:
python 3.6.8 h0371630_0 anaconda
tensorflow-gpu 1.5.0 0 anaconda
keras-applications 1.0.8 py_0 anaconda
... It`s your fault to install tensorlayer in the wrong place.
To solve your problem, activate the conda environment and then use pip to install tensorlayer.
tobimichigan
commented
5 years ago @JingqingZ & @se7enXF, I would suggest you made installation on a local machine more clearer than the current documentation indicates, I think its rather confusing. You have a good framework with a good potential. Tensorlayer is good if well maintained. I don't think the problem was my environment. Like I said, other py modules work in the same without any flaws. What I did to resolve this issue was to go on the cloud and installed tensor-layer. I am still testing your framework and will definitely post any issue that needs attention. In the nearest future, I would love to be part of your development team.
New Issue Checklist
Issue Description
[ModuleNotFoundError: No module named 'tensorlayer' after installing tensorlayer occurs when trying to run one of the git examples]
Reproducible Code
[>>> import array
@JingqingZ ..but I got this hint after running pip3 install tensorlayer "the scripts f2py, f2py3 and f2py3.7 are installed in '/home/hgh/.local/bin' which is not on PATH. Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location. The scripts imageio_download_bin and imageio_remove_bin are installed in '/home/hgh/.local/bin' which is not on PATH. Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location. The script skivi is installed in '/home/hgh/.local/bin' which is not on PATH. Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location." How do I do this, I mean add this to my path? Could this be the problem why tensorlayer cannot be found? It seems as if tensorlayer is not on the list. I was sure to activate conda env created for tensorlayer, but tensorlayer seems to be missing. Or installed elsewhere after I ran pip3 install tensorlayer.