- Added lerobot/scripts/server/gym_manipulator.py that contains all the necessary wrappers to run a gym-style env around the real robot.

- Added `lerobot/scripts/server/find_joint_limits.py` to test the min and max angles of the motion you wish the robot to explore during RL training.
- Added logic in `manipulator.py` to limit the maximum possible joint angles to allow motion within a predefined joint position range. The limits are specified in the yaml config for each robot. Checkout the so100.yaml.

Co-authored-by: Adil Zouitine <adilzouitinegm@gmail.com>

lerobot/common/robot_devices/robots/manipulator.py

@@ -45,7 +45,7 @@ def ensure_safe_goal_position(
     safe_goal_pos = present_pos + safe_diff
 
     if not torch.allclose(goal_pos, safe_goal_pos):
-        logging.warning(
+        logging.debug(
             "Relative goal position magnitude had to be clamped to be safe.\n"
             f"  requested relative goal position target: {diff}\n"
             f"    clamped relative goal position target: {safe_diff}"
@@ -464,6 +464,14 @@ class ManipulatorRobot:
             before_fwrite_t = time.perf_counter()
             goal_pos = leader_pos[name]
 
+            # If specified, clip the goal positions within predefined bounds specified in the config of the robot
+            if self.config.joint_position_relative_bounds is not None:
+                goal_pos = torch.clamp(
+                    goal_pos,
+                    self.config.joint_position_relative_bounds["min"],
+                    self.config.joint_position_relative_bounds["max"],
+                )
+
             # Cap goal position when too far away from present position.
             # Slower fps expected due to reading from the follower.
             if self.config.max_relative_target is not None:
@@ -585,6 +593,14 @@ class ManipulatorRobot:
             goal_pos = action[from_idx:to_idx]
             from_idx = to_idx
 
+            # If specified, clip the goal positions within predefined bounds specified in the config of the robot
+            if self.config.joint_position_relative_bounds is not None:
+                goal_pos = torch.clamp(
+                    goal_pos,
+                    self.config.joint_position_relative_bounds["min"],
+                    self.config.joint_position_relative_bounds["max"],
+                )
+
             # Cap goal position when too far away from present position.
             # Slower fps expected due to reading from the follower.
             if self.config.max_relative_target is not None: