From ee4a20ccdaf4516e9246b2fc0013b292456b82a9 Mon Sep 17 00:00:00 2001 From: james Date: Tue, 13 Dec 2022 11:11:37 +0800 Subject: [PATCH] Bugfix: xpu now only support single node multi-card, bkcl_comm_num should always set to 1 (#48961) --- .../fleet/meta_optimizers/graph_execution_optimizer.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/python/paddle/distributed/fleet/meta_optimizers/graph_execution_optimizer.py b/python/paddle/distributed/fleet/meta_optimizers/graph_execution_optimizer.py index 5e1cc3b1c51..fa9510aa407 100644 --- a/python/paddle/distributed/fleet/meta_optimizers/graph_execution_optimizer.py +++ b/python/paddle/distributed/fleet/meta_optimizers/graph_execution_optimizer.py @@ -137,7 +137,7 @@ class GraphExecutionOptimizer(MetaOptimizerBase): attrs={ "trainers": trainer_endpoints, "trainer_id": trainer_id, - "nccl_comm_num": build_strategy.nccl_comm_num, + "bkcl_comm_num": build_strategy.bkcl_comm_num, "use_hierarchical_allreduce": build_strategy.use_hierarchical_allreduce, "hierarchical_allreduce_inter_ranks": build_strategy.hierarchical_allreduce_inter_nranks, }, -- GitLab