diff --git a/docs/component/rl/quickstart.rst b/docs/component/rl/quickstart.rst
index 5e98e3baff..a77a99b327 100644
--- a/docs/component/rl/quickstart.rst
+++ b/docs/component/rl/quickstart.rst
@@ -54,9 +54,9 @@ QlibRL provides an example of an implementation of a single asset order executio
             # number of time indexes
             total_time: 240
             # start time index
-            default_start_time: 0
+            default_start_time_index: 0
             # end time index
-            default_end_time: 240
+            default_end_time_index: 240
             proc_data_dim: 6
         num_workers: 0
         queue_size: 20
diff --git a/examples/rl/experiment_config/backtest/config.yml b/examples/rl/experiment_config/backtest/config.yml
index 418780c2cc..0d6ad0bd1e 100644
--- a/examples/rl/experiment_config/backtest/config.yml
+++ b/examples/rl/experiment_config/backtest/config.yml
@@ -53,5 +53,5 @@ strategies:
           module_path: qlib.rl.order_execution.policy
           kwargs: 
             lr: 1.0e-4
-            weight_file: ./checkpoints/latest.pth
+            weight_file: ./checkpoints/checkpoints/latest.pth
 concurrency: 5
diff --git a/examples/rl/experiment_config/training/config.yml b/examples/rl/experiment_config/training/config.yml
index 7e50d3eee0..43f06ec914 100644
--- a/examples/rl/experiment_config/training/config.yml
+++ b/examples/rl/experiment_config/training/config.yml
@@ -32,8 +32,8 @@ data:
     order_dir: ./data/training_order_split
     data_dir: ./data/pickle_dataframe/backtest
     total_time: 240
-    default_start_time: 0
-    default_end_time: 240
+    default_start_time_index: 0
+    default_end_time_index: 240
     proc_data_dim: 6
   num_workers: 0
   queue_size: 20
diff --git a/examples/rl/simple_example.ipynb b/examples/rl/simple_example.ipynb
index 1e655ff184..7d38b3b5db 100644
--- a/examples/rl/simple_example.ipynb
+++ b/examples/rl/simple_example.ipynb
@@ -323,7 +323,7 @@
     "simulator = SimpleSimulator(100.0, NSTEPS)\n",
     "state = simulator.get_state()\n",
     "obs = [{\"obs\": state_interpreter.interpret(state)}]\n",
-    "policy_out = policy(Batch(obs))\n",
+    "policy_out = policy(Batch(obs, info=None))\n",
     "act = float(action_interpreter.interpret(state, policy_out.act))\n",
     "\n",
     "simulator.step(act)\n",
diff --git a/qlib/rl/order_execution/strategy.py b/qlib/rl/order_execution/strategy.py
index b6f5e12b24..0b84613b67 100644
--- a/qlib/rl/order_execution/strategy.py
+++ b/qlib/rl/order_execution/strategy.py
@@ -527,7 +527,7 @@ def _generate_trade_decision(self, execute_result: list | None = None) -> BaseTr
             obs_batch.append({"obs": self._state_interpreter.interpret(state)})
 
         with torch.no_grad():
-            policy_out = self._policy(Batch(obs_batch))
+            policy_out = self._policy(Batch(obs_batch, info=None))
         act = policy_out.act.numpy() if torch.is_tensor(policy_out.act) else policy_out.act
         exec_vols = [self._action_interpreter.interpret(s, a) for s, a in zip(states, act)]