MCPcopy Create free account
hub / github.com/pytorch/pytorch / DummyWork

Class DummyWork

test/distributed/test_c10d_common.py:1514–1518  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

1512
1513
1514class DummyWork(dist._Work):
1515 def wait(self, timeout=5.0):
1516 if torch.cuda.is_available():
1517 torch.cuda.current_stream().synchronize()
1518 return True
1519
1520
1521class DummyProcessGroup(dist.ProcessGroup):

Callers 7

allgatherMethod · 0.85
allreduceMethod · 0.85
barrierMethod · 0.85
broadcastMethod · 0.85
reduce_scatterMethod · 0.85
sendMethod · 0.85
recvMethod · 0.85

Calls

no outgoing calls

Tested by 7

allgatherMethod · 0.68
allreduceMethod · 0.68
barrierMethod · 0.68
broadcastMethod · 0.68
reduce_scatterMethod · 0.68
sendMethod · 0.68
recvMethod · 0.68

Used in the wild real call sites across dependent graphs

searching dependent graphs…