diff --git a/src/megatron/bridge/models/model_provider.py b/src/megatron/bridge/models/model_provider.py
index 0330e61f2f..ccf304d13d 100644
--- a/src/megatron/bridge/models/model_provider.py
+++ b/src/megatron/bridge/models/model_provider.py
@@ -677,17 +677,18 @@ def _ddp_wrap(
     else:
         DP = DistributedDataParallel
 
-    model = [
-        DP(
-            config=get_model_config(model_chunk),
-            ddp_config=ddp_config,
-            module=model_chunk,
-            # Turn off bucketing for model_chunk 2 onwards, since communication for these
-            # model chunks is overlapped with compute anyway.
-            disable_bucketing=(model_chunk_idx > 0) or overlap_param_gather_with_optimizer_step,
-        )
-        for (model_chunk_idx, model_chunk) in enumerate(model)
-    ]
+    with torch.cuda.stream(torch.cuda.Stream()):
+        model = [
+            DP(
+                config=get_model_config(model_chunk),
+                ddp_config=ddp_config,
+                module=model_chunk,
+                # Turn off bucketing for model_chunk 2 onwards, since communication for these
+                # model chunks is overlapped with compute anyway.
+                disable_bucketing=(model_chunk_idx > 0) or overlap_param_gather_with_optimizer_step,
+            )
+            for (model_chunk_idx, model_chunk) in enumerate(model)
+        ]
 
     # Broadcast params from data parallel src rank to other data parallel ranks.
     if data_parallel_random_init:
diff --git a/src/megatron/bridge/recipes/qwen/qwen3_moe.py b/src/megatron/bridge/recipes/qwen/qwen3_moe.py
index a1d6691e47..926512f8c7 100644
--- a/src/megatron/bridge/recipes/qwen/qwen3_moe.py
+++ b/src/megatron/bridge/recipes/qwen/qwen3_moe.py
@@ -225,8 +225,6 @@ def _qwen3_moe_common(
 
     if precision_config is None:
         precision_config = bf16_mixed()
-    if isinstance(precision_config, MixedPrecisionConfig):
-        precision_config.grad_reduce_in_fp32 = False
 
     # MoE-specific pipeline split configurations
     if account_for_embedding_in_pipeline_split: