You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I go to install the dependencies via cat requirements.txt | xargs -n 1 -L 1 pip install on a new conda python 3.8 environment in Ubuntu 18.04 64-bit LTS I get a number of errors especially with respect to tensorflow version and then the various geometric libraries. See screenshots below. I am working through them but if you have any leads please let me know :-)
The first error on tensorflow version was due to me using python 3.8. TF1.14 wheels don't exist for 3.8. Using 3.7 fixed the TF issue. It will help to put this in the README.md to use 3.7 :-)
But even on python 3.7 I get the errors with torch-* libraries. It seems that the wheels at the link provided have inconsistent file names and metadata causing pip to error out.
cat requirements.txt | xargs -n 1 -L 1 pip install --use-deprecated=legacy-resolve in python 3.7 works provided one pins tensorflow-gpu==1.14.0 and tensorflow-estimator==1.14.0. The deprecated resolver flag was suggested by Chi Wang at MSR.
Hi,
When I go to install the dependencies via
cat requirements.txt | xargs -n 1 -L 1 pip install
on a new conda python 3.8 environment in Ubuntu 18.04 64-bit LTS I get a number of errors especially with respect to tensorflow version and then the various geometric libraries. See screenshots below. I am working through them but if you have any leads please let me know :-)Distributor ID: Ubuntu
Description: Ubuntu 18.04.5 LTS
Release: 18.04
Codename: bionic
The text was updated successfully, but these errors were encountered: