Shukrullo Nazirjonov
Results
1
issues of
Shukrullo Nazirjonov
I am getting errors when launching the training script with `python -m torch.distributed.launch --nproc_per_node=1 --master_port=$((RANDOM)) train.py --cfg=res101_custom ` I think it is related to torch compatibility: `torch.distributed.elastic.multiprocessing.errors.ChildFailedError:` Can anyone confirm...