admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-05-27 22:49:48 +00:00
Code Issues Packages Projects Releases Wiki Activity

Refactorgym_manipulator.py using the universal pipeline (#1650)

Browse Source 
* Migrate gym_manipulator to use the pipeline
Added get_teleop_events function to capture relevant events from teleop devices unrelated to actions

* Added the capability to record a dataset

* Added the replay functionality with the pipeline

* Refactored `actor.py` to use the pipeline

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* RL works at this commit - fixed actor.py and bugs in gym_manipulator

* change folder structure to reduce the size of gym_manip

* Refactored hilserl config

* Remove dataset and mode from HilSerlEnvConfig to a GymManipulatorConfig to reduce verbose of configs during training

* format docs

* removed get_teleop_events from abc

* Refactor environment configuration and processing pipeline for GymHIL support. Removed device attribute from HILSerlRobotEnvConfig, added DummyTeleopDevice for simulation, and updated processor creation to accommodate GymHIL environments.

* Improved typing for HILRobotEnv config and GymManipulator config

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Migrated `gym_manipulator` to use a more modular structure similar to phone teleop

* Refactor gripper handling and transition processing in HIL and robot kinematic processors

- Updated gripper position handling to use a consistent key format across processors
- Improved the EEReferenceAndDelta class to handle reference joint positions.
- Added support for discrete gripper actions in the GripperVelocityToJoint processor.
- Refactored the gym manipulator to improve modularity and clarity in processing steps.

* Added delta_action_processor mapping wrapper

* Added missing file delta_action_processor and improved imports in `gym_manipulator`

* nit

* Added missing file joint_observation_processor

* Enhance processing architecture with new teleoperation processors

- Introduced `AddTeleopActionAsComplimentaryData` and `AddTeleopEventsAsInfo` for integrating teleoperator actions and events into transitions.
- Added `Torch2NumpyActionProcessor` and `Numpy2TorchActionProcessor` for seamless conversion between PyTorch tensors and NumPy arrays.
- Updated `__init__.py` to include new processors in module exports, improving modularity and clarity in the processing pipeline.
- GymHIL is now fully supported with HIL using the pipeline

* Refactor configuration structure for gym_hil integration

- Renamed sections for better readability, such as changing "Gym Wrappers Configuration" to "Processor Configuration."
- Enhanced documentation with clear examples for dataset collection and policy evaluation configurations.

* Enhance reset configuration and teleoperation event handling

- Added `terminate_on_success` parameter to `ResetConfig` and `InterventionActionProcessor` for controlling episode termination behavior upon success detection.
- Updated documentation to clarify the impact of `terminate_on_success` on data collection for reward classifier training.
- Refactored teleoperation event handling to use `TeleopEvents` constants for improved readability and maintainability across various modules.

* fix(keyboard teleop), delta action keys

* Added transform features and feature contract

* Added transform features for image crop

* Enum for TeleopEvents

* Update tranform_features delta action proc

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
Michel Aractingi
2025-08-11 11:07:55 +02:00
committed by GitHub
parent fd5d8b3d5f
commit 0053defa2e
17 changed files with 1975 additions and 2251 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/lerobot/robots/so100_follower/robot_kinematic_processor.py
+25 -7
View File
@@ -53,6 +53,9 @@ class EEReferenceAndDelta:
kinematics: RobotKinematics
end_effector_step_sizes: dict
motor_names: list[str]
use_latched_reference: bool = (
True # If True, latch reference on enable; if False, always use current pose
)
reference_ee_pose: np.ndarray | None = field(default=None, init=False, repr=False)
_prev_enabled: bool = field(default=False, init=False, repr=False)
@@ -69,7 +72,10 @@ class EEReferenceAndDelta:
"raw_joint_positions is not in complementary data and is required for EEReferenceAndDelta"
)
q = np.array([float(raw[n]) for n in self.motor_names], dtype=float)
if "reference_joint_positions" in comp:
q = comp["reference_joint_positions"]
else:
q = np.array([float(raw[n]) for n in self.motor_names], dtype=float)
# Current pose from FK on measured joints
t_curr = self.kinematics.forward_kinematics(q)
@@ -85,11 +91,12 @@ class EEReferenceAndDelta:
desired = None
if enabled:
# Latch a reference at the rising edge; also be defensive if None
if not self._prev_enabled or self.reference_ee_pose is None:
self.reference_ee_pose = t_curr.copy()
ref = self.reference_ee_pose if self.reference_ee_pose is not None else t_curr
ref = t_curr
if self.use_latched_reference:
# Latched reference mode: latch reference at the rising edge
if not self._prev_enabled or self.reference_ee_pose is None:
self.reference_ee_pose = t_curr.copy()
ref = self.reference_ee_pose if self.reference_ee_pose is not None else t_curr
delta_p = np.array(
[
@@ -100,7 +107,6 @@ class EEReferenceAndDelta:
dtype=float,
)
r_abs = Rotation.from_rotvec([wx, wy, wz]).as_matrix()
desired = np.eye(4, dtype=float)
desired[:3, :3] = ref[:3, :3] @ r_abs
desired[:3, 3] = ref[:3, 3] + delta_p
@@ -292,6 +298,8 @@ class InverseKinematicsEEToJoints:
else:
new_act[f"action.{name}.pos"] = float(q_target[i])
transition[TransitionKey.ACTION] = new_act
if not self.initial_guess_current_joints:
transition[TransitionKey.COMPLEMENTARY_DATA]["reference_joint_positions"] = q_target
return transition
def transform_features(self, features: dict[str, PolicyFeature]) -> dict[str, PolicyFeature]:
@@ -332,6 +340,7 @@ class GripperVelocityToJoint:
speed_factor: float = 20.0
clip_min: float = 0.0
clip_max: float = 100.0
discrete_gripper: bool = False
def __call__(self, transition: EnvTransition) -> EnvTransition:
obs = transition.get(TransitionKey.OBSERVATION) or {}
@@ -347,6 +356,15 @@ class GripperVelocityToJoint:
transition[TransitionKey.ACTION] = new_act
return transition
if self.discrete_gripper:
# Discrete gripper actions are in [0, 1, 2]
# 0: open, 1: close, 2: stay
# We need to shift them to [-1, 0, 1] and then scale them to clip_max
gripper_action = act.get("action.gripper", 1.0)
gripper_action = gripper_action - 1.0
gripper_action *= self.clip_max
act["action.gripper"] = gripper_action
# Get current gripper position from complementary data
raw = comp.get("raw_joint_positions") or {}
curr_pos = float(raw.get("gripper"))