diff --git a/trainer/train_agent.py b/trainer/train_agent.py
index 63a0782..4a72a11 100644
--- a/trainer/train_agent.py
+++ b/trainer/train_agent.py
@@ -471,7 +471,6 @@ if __name__ == "__main__":
     if args.use_compile == 1:
         model = torch.compile(model)
         Logger('torch.compile enabled')
-        rollout_engine.update_policy(model)
     if dist.is_initialized():
         # 同 train_ppo：RoPE buffer 各 rank 一致，每步广播纯属浪费
         model = DistributedDataParallel(model, device_ids=[local_rank], broadcast_buffers=False)
diff --git a/trainer/train_grpo.py b/trainer/train_grpo.py
index ad9aa3b..a7b01cf 100755
--- a/trainer/train_grpo.py
+++ b/trainer/train_grpo.py
@@ -312,7 +312,6 @@ if __name__ == "__main__":
     if args.use_compile == 1:
         model = torch.compile(model)
         Logger('torch.compile enabled')
-        rollout_engine.update_policy(model)
     if dist.is_initialized():
         # 同 train_ppo：RoPE buffer 各 rank 一致，每步广播纯属浪费
         model = DistributedDataParallel(model, device_ids=[local_rank], broadcast_buffers=False)
diff --git a/trainer/train_ppo.py b/trainer/train_ppo.py
index 4d762b6..b66d4b6 100644
--- a/trainer/train_ppo.py
+++ b/trainer/train_ppo.py
@@ -427,7 +427,6 @@ if __name__ == "__main__":
     if args.use_compile == 1:
         actor_model = torch.compile(actor_model)
         Logger('torch.compile enabled')
-        rollout_engine.update_policy(actor_model)
     if dist.is_initialized():
         # freqs_cos/freqs_sin 各 rank 由 config 确定性算出，默认每步广播一次纯属浪费
         actor_model = DistributedDataParallel(actor_model, device_ids=[local_rank], broadcast_buffers=False)