We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
python -m torch.distributed.launch --nproc_per_node=2 train.py
python -B inference