Add way to check if trajectory is done or max_reached

ml-agents/mlagents/trainers/bc/trainer.py

         else:
             self.episode_steps[agent_id] += len(trajectory.steps)
 
-        if trajectory.steps[-1].done:
+        if trajectory.done_reached:
             self.stats["Environment/Episode Length"].append(
                 self.episode_steps.get(agent_id, 0)
             )

ml-agents/mlagents/trainers/ppo/trainer.py

 
         value_next = self.policy.get_value_estimates(
             trajectory.next_obs,
-            trajectory.steps[-1].done and not trajectory.steps[-1].max_step,
+            trajectory.done_reached and not trajectory.done_reached,
             agent_id,
         )
 
         )
 
         # If this was a terminal trajectory, append stats and reset reward collection
-        if trajectory.steps[-1].done:
+        if trajectory.done_reached:
             self._update_end_episode_stats(agent_id)
 
     def is_ready_update(self):

ml-agents/mlagents/trainers/sac/trainer.py

             self.update_buffer, training_length=self.policy.sequence_length
         )
 
-        if trajectory.steps[-1].done:
+        if trajectory.done_reached:
             self._update_end_episode_stats(agent_id)
 
     def is_ready_update(self) -> bool:

ml-agents/mlagents/trainers/trajectory.py

             # Add the value outputs if needed
             agent_buffer_trajectory["environment_rewards"].append(exp.reward)
         return agent_buffer_trajectory
+
+    @property
+    def done_reached(self) -> bool:
+        """
+        Returns true if trajectory is terminated with a Done.
+        """
+        return self.steps[-1].done
+
+    @property
+    def max_step_reached(self) -> bool:
+        """
+        Returns true if trajectory was terminated because max steps was reached.
+        """
+        return self.steps[-1].max_step