Replies: 1 comment
-
For CUDA>=11, publishing a binary version is in our plan. However, currently you have to compile by yourself. I suggest to compile TensorFlow 2.4.1 following the document (this document is for v2.3, but v2.4 should be OK). |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hello,
As I upgraded to a RTX 3080 recently and want to use it for calculations in Lammps with Deepmd and it only supports Cuda 11.1, I want to ask how one should proceed when trying to install it. Is the latest v1.3.2 installer enough as the improvement mentioned: "Compiling with Cuda 11.0 and 11.1 #390" or do I have to do a manual installation?
I tried installing it with the run file but it seems that there is a problem with the installation as an error with Cuda 10.x is specified.
Thank you,
Alex
Beta Was this translation helpful? Give feedback.
All reactions