sharding vpp overlap bug fixer (PaddlePaddle#55365)

wz1qqx · Jul 31, 2023 · 73b2894 · 73b2894
1 parent 5715acb
commit 73b2894
Showing 1 changed file with 1 addition and 2 deletions.
diff --git a/python/paddle/distributed/fleet/meta_parallel/pipeline_parallel.py b/python/paddle/distributed/fleet/meta_parallel/pipeline_parallel.py
@@ -298,12 +298,11 @@ def register_allreduce_overlap_hook(self, model, comm_group, acc_steps, dp):
 
         act = HOOK_ACTION.ALL_REDUCE if dp else HOOK_ACTION.REDUCE
 
-        fused_parameter_group = {}
-
         for model in models:
             # For virtual pipeline. Will separate parameters in different chunk into
             # different groups to get the best performance.
 
+            fused_parameter_group = {}
             parameter_list = [
                 p for p in model.parameters() if not p.stop_gradient
             ]