CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutPoliciesSign UpSign In
ai-forever

Real-time collaboration for Jupyter Notebooks, Linux Terminals, LaTeX, VS Code, R IDE, and more,
all in one place. Commercial Alternative to JupyterHub.

GitHub Repository: ai-forever/sber-swap
Path: blob/main/apex/tests/distributed/synced_batchnorm/unit_test.sh
Views: 794
1
python python_single_gpu_unit_test.py
2
python single_gpu_unit_test.py
3
python test_batchnorm1d.py
4
python -m torch.distributed.launch --nproc_per_node=2 two_gpu_unit_test.py
5
python -m torch.distributed.launch --nproc_per_node=2 two_gpu_unit_test.py --fp16
6
python -m torch.distributed.launch --nproc_per_node=2 two_gpu_test_different_batch_size.py --apex
7
#beware, you need a system with at least 4 gpus to test group_size<world_size
8
#python -m torch.distributed.launch --nproc_per_node=4 test_groups.py --group_size=2
9
10