Unanswered
Clearml (Remote Execution) Sometimes Doesn'T "Pick-Up" Gpu. After I Rerun The Task It Picks It Up. Seems Random, Doesn'T Happen Too Often (Maybe Once In 30-40 Times) And I Cannot Seem To Detect Any Pattern. Did Anyone Else Notice This?
Agents Are Vms On G
"Executing: ['docker', 'run', '-t', '--gpus', '"device=0"'" - so the container is executed with --gpus.
However, torch.cuda.is_available() returns False.
37 Views
0
Answers
3 months ago
3 months ago