diff --git a/verl/workers/actor/dp_actor.py b/verl/workers/actor/dp_actor.py
index ff762b6..6f384b0 100644
--- a/verl/workers/actor/dp_actor.py
+++ b/verl/workers/actor/dp_actor.py
@@ -142,7 +142,7 @@ def _make_minibatch_iterator(self, data: DataProto) -> Iterable[DataProto]:
         data = data.select(batch_keys=select_keys)
         return data.make_iterator(mini_batch_size=self.config.ppo_mini_batch_size,
                                   epochs=self.config.ppo_epochs,
-                                  dataloader_kwargs={'shuffle': self.config.shuffle})
+                                  dataloader_kwargs={'shuffle': False}) # TODO: hardcode to False
 
     def _optimizer_step(self):
         assert self.config.grad_clip is not None
diff --git a/verl/workers/critic/dp_critic.py b/verl/workers/critic/dp_critic.py
index 3d17b0a..d9e7ffa 100644
--- a/verl/workers/critic/dp_critic.py
+++ b/verl/workers/critic/dp_critic.py
@@ -104,7 +104,7 @@ def _make_minibatch_iterator(self, data: DataProto) -> Iterable[DataProto]:
         data = data.select(batch_keys=select_keys)
         return data.make_iterator(mini_batch_size=self.config.ppo_mini_batch_size,
                                   epochs=self.config.ppo_epochs,
-                                  dataloader_kwargs={'shuffle': self.config.shuffle})
+                                  dataloader_kwargs={'shuffle': False}) # TODO: hardcode to False
 
     def _optimizer_step(self):
         assert self.config.grad_clip is not None