Unanswered
Hi, I Shifted My Clearml Setup To An On-Premise Disconnected Env, Which Has A Pip Repo Setup. I Noted This Warning,
I agree with Martin.B, it appears to be a CUDA mismatch. The version of torch is trying to use cuda 10.2 but you haveagent.default_docker.image = nvidia/cuda:10.1-runtime-ubuntu18.04
that should probably beagent.default_docker.image = nvidia/cuda:10.2-runtime-ubuntu18.04
163 Views
0
Answers
3 years ago
one year ago