Searched refs:all_reduce_coalesced (Results 1 – 17 of 17) sorted by relevance
| /external/pytorch/test/distributed/_tensor/debug/ |
| D | test_comm_mode.py | 80 return funcol.all_reduce_coalesced([out], "sum", world_pg)
90 self.assertEqual(comm_counts[c10d_functional.all_reduce_coalesced], 1)
194 dist.all_reduce_coalesced(inp)
|
| /external/pytorch/test/distributed/ |
| D | test_c10d_functional_native.py | 15 all_reduce_coalesced,
139 outputs = torch.ops._c10d_functional.all_reduce_coalesced(
150 outputs = all_reduce_coalesced(
544 ar0 = funcol.all_reduce_coalesced(bufs, "avg", "0")
547 ar1 = funcol.all_reduce_coalesced(args, "avg", "0")
|
| D | test_multi_threaded_pg.py | 297 dist.all_reduce_coalesced([t0, t1])
|
| D | test_c10d_common.py | 1311 dist.all_reduce_coalesced(tensor_list_h)
1908 dist.all_reduce_coalesced(tensors, dist.ReduceOp.SUM)
|
| D | test_functional_api.py | 278 res = ft_c.all_reduce_coalesced([t0, t1], "sum", mesh)
|
| D | test_c10d_nccl.py | 2728 torch.distributed.all_reduce_coalesced(tensors, group=process_group)
2756 torch.distributed.all_reduce_coalesced(tensors, group=process_group)
|
| D | test_c10d_gloo.py | 597 futs = [c10d.all_reduce_coalesced(x, async_op=True) for x in xs]
|
| /external/pytorch/test/inductor/ |
| D | test_snode_runtime.py | 236 rs = c10d.all_reduce_coalesced(x, "sum", "", self.RANKS, self.WORLD_SIZE)
265 rs = _c10d.all_reduce_coalesced(x, "sum", "0")
|
| /external/pytorch/torch/distributed/ |
| D | _functional_collectives_impl.py | 34 return torch.ops._c10d_functional.all_reduce_coalesced(
|
| D | _functional_collectives.py | 332 def all_reduce_coalesced( function
352 tensor_list = torch.ops._c10d_functional.all_reduce_coalesced( # type: ignore[attr-defined]
|
| D | distributed_c10d.py | 2516 def all_reduce_coalesced(tensors, op=ReduceOp.SUM, group=None, async_op=False): function
4968 all_reduce_coalesced,
|
| /external/pytorch/torch/distributed/tensor/debug/ |
| D | _comm_mode.py | 35 funcol_native.all_reduce_coalesced: funcol_py.all_reduce_coalesced,
|
| /external/pytorch/torch/_inductor/fx_passes/ |
| D | ddp_fusion.py | 268 graph, torch.ops._c10d_functional.all_reduce_coalesced.default, args, kwargs
543 torch.ops._c10d_functional.all_reduce_coalesced.default,
|
| D | reinplace.py | 344 c10d_functional.all_reduce_coalesced.default: InplaceableOp(
|
| /external/pytorch/torch/csrc/distributed/c10d/ |
| D | Functional.cpp | 144 std::vector<at::Tensor> all_reduce_coalesced( in all_reduce_coalesced() function
325 c10::DispatchKey::CompositeExplicitAutograd, ::all_reduce_coalesced), in TORCH_LIBRARY()
|
| /external/pytorch/torch/testing/_internal/distributed/ |
| D | distributed_test.py | 3027 dist.all_reduce_coalesced(
3068 dist.all_reduce_coalesced,
|
| /external/pytorch/torch/_inductor/ |
| D | lowering.py | 6331 @register_lowering(_c10d_functional.all_reduce_coalesced)
|