E5E7 [Distributed][CI] Rework continuous TestCase by kwen2501 · Pull Request #153653 · pytorch/pytorch · GitHub
[go: up one dir, main page]
Skip to content

[Distributed][CI] Rework continuous TestCase #153653

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
kwen2501 wants to merge 8 commits into from
Closed

[Distributed][CI] Rework continuous TestCase #153653

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
8f9d471
Update
kwen2501 May 15, 2025
141207a
Update
kwen2501 May 16, 2025
6f6d7de
Update
kwen2501 May 16, 2025
6ff0db8
Update
kwen2501 May 17, 2025
de55afe
Update
kwen2501 May 19, 2025
0ecc3d3
Update
kwen2501 May 23, 2025
4b820a8
Update
kwen2501 May 24, 2025
95f6d5d
Update
kwen2501 May 24, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions .lintrunner.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -990,6 +990,9 @@ exclude_patterns = [
'test/distributed/optim/test_apply_optimizer_in_backward.py',
'test/distributed/optim/test_named_optimizer.py',
'test/distributed/test_c10d_spawn.py',
'test/distributed/test_c10d_ops_nccl.py',
'test/distributed/test_symmetric_memory.py',
'test/distributed/test_nvshmem.py',
'test/distributed/test_collective_utils.py',
'test/distributions/test_distributions.py',
'test/inductor/test_aot_inductor_utils.py',
Expand Down
39 changes: 19 additions & 20 deletions test/distributed/test_c10d_ops_nccl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1045,23 +1045,22 @@ def allgather_base(output_t, input_t):

if __name__ == "__main__":
if not torch.cuda.is_available():
sys.exit(TEST_SKIPS["no_cuda"].exit_code)

rank = int(os.getenv("RANK", -1))
world_size = int(os.getenv("WORLD_SIZE", -1))

if world_size == -1: # Not set by external launcher
world_size = torch.cuda.device_count()

if rank != -1:
# Launched with torchrun or other multi-proc launchers. Directly run the test.
ProcessGroupNCCLOpTest.run_rank(rank, world_size)
else:
# Launched as a single process. Spawn subprocess to run the tests.
# Also need a rendezvous file for `init_process_group` purpose.
rdvz_file = tempfile.NamedTemporaryFile(delete=False).name
torch.multiprocessing.spawn(
ProcessGroupNCCLOpTest.run_rank,
nprocs=world_size,
args=(world_size, rdvz_file),
)
os._exit(TEST_SKIPS["no_cuda"].exit_code)

# Use device count as world size
world_size = torch.cuda.device_count()
# Also need a rendezvous file for `init_process_group` purpose.
rdvz_file = tempfile.NamedTemporaryFile(delete=False).name
# Spawn subprocess to run the tests.
# `run_tests()` will be called under `run_rank`
torch.multiprocessing.spawn(
MultiProcContinousTest.run_rank, # entry point
nprocs=world_size,
args=(world_size, rdvz_file),
)

# Clear up the rendezvous file
try:
os.remove(rdvz_file)
except OSError:
pass
Loading
Loading
0