Abhimanyu Chadha
Abhimanyu Chadha
@maxhgerlach Any luck with this one? Opened a new ticket as you'd mentioned the previous one had a different OS and platform.
So, Conda is just for some dependencies. And the result is the same without `sudo pip` too. I shall post the full output in the next command.
[horovod_install_log.txt](https://github.com/horovod/horovod/files/8941346/horovod_install_log.txt) [horovod_import_error_log.txt](https://github.com/horovod/horovod/files/8941349/horovod_import_error_log.txt)
Thank you for the quick reply. For the second error, I am on a beta version of MacOS which was released less than a week ago. So I can look...
@adamjstewart As per the developers at TF, they did say that a new build should fix this issue but when they're going to push that new build, I have no...
@maxhgerlach Any luck with this on your side? Haven't heard from the developers at TF in a while.
@maxhgerlach This issue has been fixed in Tensorflow 2.10
@maxhgerlach In continuation to the previous bug, now I am able to import everything but I can't find any examples for running or detecting the M1 chips in the horovod...
@AdamStormhardtGH Which python version are you running on your M1? I am running 3.9 and am unable to install bpy.