hangX/lerobot
1
0
Fork 0
forked from tangger/lerobot
Code Pull Requests Activity

Fix linter issue

Browse Source
This commit is contained in:
AdilZouitine
2025-04-22 10:37:08 +02:00
parent 0030ff3f74
commit c5845ee203
14 changed files with 64 additions and 1667 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/12_train_hilserl_classifier.md
View File

@@ -42,7 +42,7 @@ Replace the `dataset_repo_id` field with the identifier for your dataset, which
dataset_repo_id: "my_dataset_repo_id"
## Typical logs and metrics
```
When you start the training process, you will first see your full configuration being printed in the terminal. You can check it to make sure that you config it correctly and your config is not overrided by other files. The final configuration will also be saved with the checkpoint.
When you start the training process, you will first see your full configuration being printed in the terminal. You can check it to make sure that you config it correctly and your config is not overriden by other files. The final configuration will also be saved with the checkpoint.
After that, you will see training log like this one:

4
lerobot/common/robot_devices/control_utils.py
View File

@@ -262,8 +262,6 @@ def control_loop(
while timestamp < control_time_s:
start_loop_t = time.perf_counter()
current_joint_positions = robot.follower_arms["main"].read("Present_Position")
if teleoperate:
observation, action = robot.teleop_step(record_data=True)
else:
@@ -326,7 +324,7 @@ def reset_follower_position(robot: Robot, target_position):
current_position = robot.follower_arms["main"].read("Present_Position")
trajectory = torch.from_numpy(
np.linspace(current_position, target_position, 50)
) # NOTE: 30 is just an aribtrary number
) # NOTE: 30 is just an arbitrary number
for pose in trajectory:
robot.send_action(pose)
busy_wait(0.015)

412
lerobot/scripts/eval_on_robot.py
View File

@@ -1,412 +0,0 @@
#!/usr/bin/env python
# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Evaluate a policy by running rollouts on the real robot and computing metrics.
Usage examples: evaluate a checkpoint from the LeRobot training script for 10 episodes.
```
python lerobot/scripts/eval_on_robot.py \
-p outputs/train/model/checkpoints/005000/pretrained_model \
eval.n_episodes=10
```
Test reward classifier with teleoperation (you need to press space to take over)
```
python lerobot/scripts/eval_on_robot.py \
--robot-path lerobot/configs/robot/so100.yaml \
--reward-classifier-pretrained-path outputs/classifier/checkpoints/best/pretrained_model \
--reward-classifier-config-file lerobot/configs/policy/hilserl_classifier.yaml \
--display-cameras 1
```
**NOTE** (michel-aractingi): This script is incomplete and it is being prepared
for running training on the real robot.
"""
import argparse
import logging
import time
import cv2
import numpy as np
import torch
from tqdm import trange
from lerobot.common.policies.policy_protocol import Policy
from lerobot.common.robot_devices.control_utils import (
busy_wait,
is_headless,
reset_follower_position,
)
from lerobot.common.robot_devices.robots.factory import Robot, make_robot
from lerobot.common.utils.utils import (
init_hydra_config,
init_logging,
log_say,
)
def get_classifier(pretrained_path, config_path):
if pretrained_path is None or config_path is None:
return
from lerobot.common.policies.factory import _policy_cfg_from_hydra_cfg
from lerobot.common.policies.hilserl.classifier.configuration_classifier import (
ClassifierConfig,
)
from lerobot.common.policies.hilserl.classifier.modeling_classifier import (
Classifier,
)
cfg = init_hydra_config(config_path)
classifier_config = _policy_cfg_from_hydra_cfg(ClassifierConfig, cfg)
classifier_config.num_cameras = len(cfg.training.image_keys) # TODO automate these paths
model = Classifier(classifier_config)
model.load_state_dict(Classifier.from_pretrained(pretrained_path).state_dict())
model = model.to("mps")
return model
def rollout(
robot: Robot,
policy: Policy,
reward_classifier,
fps: int,
control_time_s: float = 20,
use_amp: bool = True,
display_cameras: bool = False,
) -> dict:
"""Run a batched policy rollout on the real robot.
The return dictionary contains:
"robot": A a dictionary of (batch, sequence + 1, *) tensors mapped to observation
keys. NOTE the that this has an extra sequence element relative to the other keys in the
dictionary. This is because an extra observation is included for after the environment is
terminated or truncated.
"action": A (batch, sequence, action_dim) tensor of actions applied based on the observations (not
including the last observations).
"reward": A (batch, sequence) tensor of rewards received for applying the actions.
"success": A (batch, sequence) tensor of success conditions (the only time this can be True is upon
environment termination/truncation).
"done": A (batch, sequence) tensor of **cumulative** done conditions. For any given batch element,
the first True is followed by True's all the way till the end. This can be used for masking
extraneous elements from the sequences above.
Args:
robot: The robot class that defines the interface with the real robot.
policy: The policy. Must be a PyTorch nn module.
Returns:
The dictionary described above.
"""
# TODO (michel-aractingi): Infer the device from policy parameters when policy is added
# assert isinstance(policy, nn.Module), "Policy must be a PyTorch nn module."
# device = get_device_from_parameters(policy)
# define keyboard listener
listener, events = init_keyboard_listener()
# Reset the policy. TODO (michel-aractingi) add real policy evaluation once the code is ready.
# policy.reset()
# NOTE: sorting to make sure the key sequence is the same during training and testing.
observation = robot.capture_observation()
image_keys = [key for key in observation if "image" in key]
image_keys.sort()
all_actions = []
all_rewards = []
all_successes = []
start_episode_t = time.perf_counter()
init_pos = robot.follower_arms["main"].read("Present_Position")
timestamp = 0.0
while timestamp < control_time_s:
start_loop_t = time.perf_counter()
# Apply the next action.
while events["pause_policy"] and not events["human_intervention_step"]:
busy_wait(0.5)
if events["human_intervention_step"]:
# take over the robot's actions
observation, action = robot.teleop_step(record_data=True)
action = action["action"] # teleop step returns torch tensors but in a dict
else:
# explore with policy
with torch.inference_mode():
# TODO (michel-aractingi) replace this part with policy (predict_action)
action = robot.follower_arms["main"].read("Present_Position")
action = torch.from_numpy(action)
robot.send_action(action)
# action = predict_action(observation, policy, device, use_amp)
observation = robot.capture_observation()
images = []
for key in image_keys:
if display_cameras:
cv2.imshow(key, cv2.cvtColor(observation[key].numpy(), cv2.COLOR_RGB2BGR))
cv2.waitKey(1)
images.append(observation[key].to("mps"))
reward = reward_classifier.predict_reward(images) if reward_classifier is not None else 0.0
all_rewards.append(reward)
# print("REWARD : ", reward)
all_actions.append(action)
all_successes.append(torch.tensor([False]))
dt_s = time.perf_counter() - start_loop_t
busy_wait(1 / fps - dt_s)
timestamp = time.perf_counter() - start_episode_t
if events["exit_early"]:
events["exit_early"] = False
events["human_intervention_step"] = False
events["pause_policy"] = False
break
reset_follower_position(robot, target_position=init_pos)
dones = torch.tensor([False] * len(all_actions))
dones[-1] = True
# Stack the sequence along the first dimension so that we have (batch, sequence, *) tensors.
ret = {
"action": torch.stack(all_actions, dim=1),
"next.reward": torch.stack(all_rewards, dim=1),
"next.success": torch.stack(all_successes, dim=1),
"done": dones,
}
listener.stop()
return ret
def eval_policy(
robot: Robot,
policy: torch.nn.Module,
fps: float,
n_episodes: int,
control_time_s: int = 20,
use_amp: bool = True,
display_cameras: bool = False,
reward_classifier_pretrained_path: str | None = None,
reward_classifier_config_file: str | None = None,
) -> dict:
"""
Args:
env: The batch of environments.
policy: The policy.
n_episodes: The number of episodes to evaluate.
Returns:
Dictionary with metrics and data regarding the rollouts.
"""
# TODO (michel-aractingi) comment this out for testing with a fixed policy
# assert isinstance(policy, Policy)
# policy.eval()
sum_rewards = []
max_rewards = []
successes = []
rollouts = []
start_eval = time.perf_counter()
progbar = trange(n_episodes, desc="Evaluating policy on real robot")
reward_classifier = get_classifier(reward_classifier_pretrained_path, reward_classifier_config_file)
for _ in progbar:
rollout_data = rollout(
robot,
policy,
reward_classifier,
fps,
control_time_s,
use_amp,
display_cameras,
)
rollouts.append(rollout_data)
sum_rewards.append(sum(rollout_data["next.reward"]))
max_rewards.append(max(rollout_data["next.reward"]))
successes.append(rollout_data["next.success"][-1])
info = {
"per_episode": [
{
"episode_ix": i,
"sum_reward": sum_reward,
"max_reward": max_reward,
"pc_success": success * 100,
}
for i, (sum_reward, max_reward, success) in enumerate(
zip(
sum_rewards[:n_episodes],
max_rewards[:n_episodes],
successes[:n_episodes],
strict=False,
)
)
],
"aggregated": {
"avg_sum_reward": float(np.nanmean(torch.cat(sum_rewards[:n_episodes]))),
"avg_max_reward": float(np.nanmean(torch.cat(max_rewards[:n_episodes]))),
"pc_success": float(np.nanmean(torch.cat(successes[:n_episodes])) * 100),
"eval_s": time.time() - start_eval,
"eval_ep_s": (time.time() - start_eval) / n_episodes,
},
}
if robot.is_connected:
robot.disconnect()
return info
def init_keyboard_listener():
# Allow to exit early while recording an episode or resetting the environment,
# by tapping the right arrow key '->'. This might require a sudo permission
# to allow your terminal to monitor keyboard events.
events = {}
events["exit_early"] = False
events["rerecord_episode"] = False
events["pause_policy"] = False
events["human_intervention_step"] = False
if is_headless():
logging.warning(
"Headless environment detected. On-screen cameras display and keyboard inputs will not be available."
)
listener = None
return listener, events
# Only import pynput if not in a headless environment
from pynput import keyboard
def on_press(key):
try:
if key == keyboard.Key.right:
print("Right arrow key pressed. Exiting loop...")
events["exit_early"] = True
elif key == keyboard.Key.left:
print("Left arrow key pressed. Exiting loop and rerecord the last episode...")
events["rerecord_episode"] = True
events["exit_early"] = True
elif key == keyboard.Key.space:
# check if first space press then pause the policy for the user to get ready
# if second space press then the user is ready to start intervention
if not events["pause_policy"]:
print(
"Space key pressed. Human intervention required.\n"
"Place the leader in similar pose to the follower and press space again."
)
events["pause_policy"] = True
log_say(
"Human intervention stage. Get ready to take over.",
play_sounds=True,
)
else:
events["human_intervention_step"] = True
print("Space key pressed. Human intervention starting.")
log_say("Starting human intervention.", play_sounds=True)
except Exception as e:
print(f"Error handling key press: {e}")
listener = keyboard.Listener(on_press=on_press)
listener.start()
return listener, events
if __name__ == "__main__":
init_logging()
parser = argparse.ArgumentParser(
description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter
)
group = parser.add_mutually_exclusive_group(required=True)
group.add_argument(
"--robot-path",
type=str,
default="lerobot/configs/robot/koch.yaml",
help="Path to robot yaml file used to instantiate the robot using `make_robot` factory function.",
)
group.add_argument(
"--robot-overrides",
type=str,
nargs="*",
help="Any key=value arguments to override config values (use dots for.nested=overrides)",
)
group.add_argument(
"-p",
"--pretrained-policy-name-or-path",
help=(
"Either the repo ID of a model hosted on the Hub or a path to a directory containing weights "
"saved using `Policy.save_pretrained`. If not provided, the policy is initialized from scratch "
"(useful for debugging). This argument is mutually exclusive with `--config`."
),
)
group.add_argument(
"--config",
help=(
"Path to a yaml config you want to use for initializing a policy from scratch (useful for "
"debugging). This argument is mutually exclusive with `--pretrained-policy-name-or-path` (`-p`)."
),
)
parser.add_argument("--revision", help="Optionally provide the Hugging Face Hub revision ID.")
parser.add_argument(
"--out-dir",
help=(
"Where to save the evaluation outputs. If not provided, outputs are saved in "
"outputs/eval/{timestamp}_{env_name}_{policy_name}"
),
)
parser.add_argument(
"--display-cameras",
help=("Whether to display the camera feed while the rollout is happening"),
)
parser.add_argument(
"--reward-classifier-pretrained-path",
type=str,
default=None,
help="Path to the pretrained classifier weights.",
)
parser.add_argument(
"--reward-classifier-config-file",
type=str,
default=None,
help="Path to a yaml config file that is necessary to build the reward classifier model.",
)
args = parser.parse_args()
robot_cfg = init_hydra_config(args.robot_path, args.robot_overrides)
robot = make_robot(robot_cfg)
if not robot.is_connected:
robot.connect()
eval_policy(
robot,
None,
fps=40,
n_episodes=2,
control_time_s=100,
display_cameras=args.display_cameras,
reward_classifier_config_file=args.reward_classifier_config_file,
reward_classifier_pretrained_path=args.reward_classifier_pretrained_path,
)

4
lerobot/scripts/server/actor_server.py
View File

@@ -224,7 +224,7 @@ def act_with_policy(
logging.info("make_policy")
### Instantiate the policy in both the actor and learner processes
### To avoid sending a SACPolicy object through the port, we create a policy intance
### To avoid sending a SACPolicy object through the port, we create a policy instance
### on both sides, the learner sends the updated parameters every n steps to update the actor's parameters
# TODO: At some point we should just need make sac policy
policy: SACPolicy = make_policy(
@@ -278,7 +278,7 @@ def act_with_policy(
# Increment total steps counter for intervention rate
episode_total_steps += 1
# NOTE: We overide the action if the intervention is True, because the action applied is the intervention action
# NOTE: We override the action if the intervention is True, because the action applied is the intervention action
if "is_intervention" in info and info["is_intervention"]:
# TODO: Check the shape
# NOTE: The action space for demonstration before hand is with the full action space

2
lerobot/scripts/server/crop_dataset_roi.py
View File

@@ -269,7 +269,7 @@ if __name__ == "__main__":
new_repo_id = args.repo_id + "_cropped_resized"
new_dataset_root = Path(str(dataset.root) + "_cropped_resized")
croped_resized_dataset = convert_lerobot_dataset_to_cropper_lerobot_dataset(
cropped_resized_dataset = convert_lerobot_dataset_to_cropper_lerobot_dataset(
original_dataset=dataset,
crop_params_dict=rois,
new_repo_id=new_repo_id,

73
lerobot/scripts/server/end_effector_control_utils.py
View File

@@ -262,7 +262,7 @@ class GamepadController(InputController):
elif event.button == 6:
self.close_gripper_command = True
# LT button (7) for openning gripper
# LT button (7) for opening gripper
elif event.button == 7:
self.open_gripper_command = True
@@ -421,45 +421,44 @@ class GamepadControllerHID(InputController):
try:
# Read data from the gamepad
data = self.device.read(64)
if data:
# Interpret gamepad data - this will vary by controller model
# These offsets are for the Logitech RumblePad 2
if len(data) >= 8:
# Normalize joystick values from 0-255 to -1.0-1.0
self.left_x = (data[1] - 128) / 128.0
self.left_y = (data[2] - 128) / 128.0
self.right_x = (data[3] - 128) / 128.0
self.right_y = (data[4] - 128) / 128.0
# Interpret gamepad data - this will vary by controller model
# These offsets are for the Logitech RumblePad 2
if data and len(data) >= 8:
# Normalize joystick values from 0-255 to -1.0-1.0
self.left_x = (data[1] - 128) / 128.0
self.left_y = (data[2] - 128) / 128.0
self.right_x = (data[3] - 128) / 128.0
self.right_y = (data[4] - 128) / 128.0
# Apply deadzone
self.left_x = 0 if abs(self.left_x) < self.deadzone else self.left_x
self.left_y = 0 if abs(self.left_y) < self.deadzone else self.left_y
self.right_x = 0 if abs(self.right_x) < self.deadzone else self.right_x
self.right_y = 0 if abs(self.right_y) < self.deadzone else self.right_y
# Apply deadzone
self.left_x = 0 if abs(self.left_x) < self.deadzone else self.left_x
self.left_y = 0 if abs(self.left_y) < self.deadzone else self.left_y
self.right_x = 0 if abs(self.right_x) < self.deadzone else self.right_x
self.right_y = 0 if abs(self.right_y) < self.deadzone else self.right_y
# Parse button states (byte 5 in the Logitech RumblePad 2)
buttons = data[5]
# Parse button states (byte 5 in the Logitech RumblePad 2)
buttons = data[5]
# Check if RB is pressed then the intervention flag should be set
self.intervention_flag = data[6] in [2, 6, 10, 14]
# Check if RB is pressed then the intervention flag should be set
self.intervention_flag = data[6] in [2, 6, 10, 14]
# Check if RT is pressed
self.open_gripper_command = data[6] in [8, 10, 12]
# Check if RT is pressed
self.open_gripper_command = data[6] in [8, 10, 12]
# Check if LT is pressed
self.close_gripper_command = data[6] in [4, 6, 12]
# Check if LT is pressed
self.close_gripper_command = data[6] in [4, 6, 12]
# Check if Y/Triangle button (bit 7) is pressed for saving
# Check if X/Square button (bit 5) is pressed for failure
# Check if A/Cross button (bit 4) is pressed for rerecording
if buttons & 1 << 7:
self.episode_end_status = "success"
elif buttons & 1 << 5:
self.episode_end_status = "failure"
elif buttons & 1 << 4:
self.episode_end_status = "rerecord_episode"
else:
self.episode_end_status = None
# Check if Y/Triangle button (bit 7) is pressed for saving
# Check if X/Square button (bit 5) is pressed for failure
# Check if A/Cross button (bit 4) is pressed for rerecording
if buttons & 1 << 7:
self.episode_end_status = "success"
elif buttons & 1 << 5:
self.episode_end_status = "failure"
elif buttons & 1 << 4:
self.episode_end_status = "rerecord_episode"
else:
self.episode_end_status = None
except OSError as e:
logging.error(f"Error reading from gamepad: {e}")
@@ -618,7 +617,7 @@ def teleoperate_delta_inverse_kinematics(robot, controller, fps=10, bounds=None,
# Process input events
controller.update()
# Get currrent robot state
# Get current robot state
joint_positions = robot.follower_arms["main"].read("Present_Position")
current_ee_pos = kinematics.fk_gripper_tip(joint_positions)
@@ -635,7 +634,7 @@ def teleoperate_delta_inverse_kinematics(robot, controller, fps=10, bounds=None,
desired_ee_pos[:3, 3] = np.clip(desired_ee_pos[:3, 3], bounds["min"], bounds["max"])
# Only send commands if there's actual movement
if any([abs(v) > 0.001 for v in [delta_x, delta_y, delta_z]]):
if any(abs(v) > 0.001 for v in [delta_x, delta_y, delta_z]):
# Compute joint targets via inverse kinematics
target_joint_state = kinematics.ik(joint_positions, desired_ee_pos, position_only=True)
@@ -678,7 +677,7 @@ def teleoperate_gym_env(env, controller, fps: int = 30):
action = np.array([delta_x, delta_y, delta_z])
# Skip if no movement
if any([abs(v) > 0.001 for v in [delta_x, delta_y, delta_z]]):
if any(abs(v) > 0.001 for v in [delta_x, delta_y, delta_z]):
# Step the environment - pass action as a tensor with intervention flag
action_tensor = torch.from_numpy(action.astype(np.float32))
obs, reward, terminated, truncated, info = env.step((action_tensor, False))

25
lerobot/scripts/server/gym_manipulator.py
View File

@@ -106,7 +106,7 @@ class HILSerlRobotEnv(gym.Env):
- The action space is defined as a Tuple where:
• The first element is a Box space representing joint position commands. It is defined as relative (delta)
or absolute, based on the configuration.
• ThE SECONd element is a Discrete space (with 2 values) serving as a flag for intervention (teleoperation).
• The second element is a Discrete space (with 2 values) serving as a flag for intervention (teleoperation).
"""
example_obs = self.robot.capture_observation()
@@ -384,7 +384,7 @@ class ActionRepeatWrapper(gym.Wrapper):
class RewardWrapper(gym.Wrapper):
def __init__(self, env, reward_classifier, device: torch.device = "cuda"):
"""
Wrapper to add reward prediction to the environment, it use a trained classifer.
Wrapper to add reward prediction to the environment, it use a trained classifier.
cfg.
env: The environment to wrap
@@ -414,7 +414,7 @@ class RewardWrapper(gym.Wrapper):
if self.reward_classifier is not None
else 0.0
)
info["Reward classifer frequency"] = 1 / (time.perf_counter() - start_time)
info["Reward classifier frequency"] = 1 / (time.perf_counter() - start_time)
if success == 1.0:
terminated = True
@@ -784,11 +784,11 @@ class ResetWrapper(gym.Wrapper):
while time.perf_counter() - start_time < self.reset_time_s:
self.robot.teleop_step()
log_say("Manual reseting of the environment done.", play_sounds=True)
log_say("Manual reset of the environment done.", play_sounds=True)
return super().reset(seed=seed, options=options)
class BatchCompitableWrapper(gym.ObservationWrapper):
class BatchCompatibleWrapper(gym.ObservationWrapper):
def __init__(self, env):
super().__init__(env)
@@ -823,10 +823,7 @@ class GripperPenaltyWrapper(gym.RewardWrapper):
def step(self, action):
self.last_gripper_state = self.unwrapped.robot.follower_arms["main"].read("Present_Position")[-1]
if isinstance(action, tuple):
gripper_action = action[0][-1]
else:
gripper_action = action[-1]
gripper_action = action[0][-1] if isinstance(action, tuple) else action[-1]
obs, reward, terminated, truncated, info = self.env.step(action)
gripper_penalty = self.reward(reward, gripper_action)
@@ -1279,7 +1276,7 @@ def make_robot_env(cfg) -> gym.vector.VectorEnv:
)
if cfg.wrapper.ee_action_space_params is None and cfg.wrapper.joint_masking_action_space is not None:
env = JointMaskingActionSpace(env=env, mask=cfg.wrapper.joint_masking_action_space)
env = BatchCompitableWrapper(env=env)
env = BatchCompatibleWrapper(env=env)
return env
@@ -1492,7 +1489,7 @@ def main(cfg: EnvConfig):
alpha = 1.0
num_episode = 0
sucesses = []
successes = []
while num_episode < 20:
start_loop_s = time.perf_counter()
# Sample a new random action from the robot's action space.
@@ -1503,15 +1500,15 @@ def main(cfg: EnvConfig):
# Execute the step: wrap the NumPy action in a torch tensor.
obs, reward, terminated, truncated, info = env.step((torch.from_numpy(smoothed_action), False))
if terminated or truncated:
sucesses.append(reward)
successes.append(reward)
env.reset()
num_episode += 1
dt_s = time.perf_counter() - start_loop_s
busy_wait(1 / cfg.fps - dt_s)
logging.info(f"Success after 20 steps {sucesses}")
logging.info(f"success rate {sum(sucesses) / len(sucesses)}")
logging.info(f"Success after 20 steps {successes}")
logging.info(f"success rate {sum(successes) / len(successes)}")
if __name__ == "__main__":

8
lerobot/scripts/server/learner_server.py
View File

@@ -319,7 +319,7 @@ def add_actor_information_and_train(
batch_size: int = batch_size // 2 # We will sample from both replay buffer
logging.info("Starting learner thread")
interaction_message, transition = None, None
interaction_message = None
optimization_step = resume_optimization_step if resume_optimization_step is not None else 0
interaction_step_shift = resume_interaction_step if resume_interaction_step is not None else 0
@@ -654,7 +654,7 @@ def start_learner_server(
shutdown_event.wait()
logging.info("[LEARNER] Stopping gRPC server...")
server.stop(learner_service.STUTDOWN_TIMEOUT)
server.stop(learner_service.SHUTDOWN_TIMEOUT)
logging.info("[LEARNER] gRPC server stopped")
@@ -719,7 +719,7 @@ def save_training_checkpoint(
# Update the "last" symlink
update_last_checkpoint(checkpoint_dir)
# TODO : temporarly save replay buffer here, remove later when on the robot
# TODO : temporary save replay buffer here, remove later when on the robot
# We want to control this with the keyboard inputs
dataset_dir = os.path.join(cfg.output_dir, "dataset")
if os.path.exists(dataset_dir) and os.path.isdir(dataset_dir):
@@ -889,7 +889,7 @@ def load_training_state(
training_state_path = os.path.join(checkpoint_dir, TRAINING_STATE_DIR, "training_state.pt")
interaction_step = 0
if os.path.exists(training_state_path):
training_state = torch.load(training_state_path, weights_only=False)
training_state = torch.load(training_state_path, weights_only=False) # nosec B614: Safe usage of torch.load
interaction_step = training_state.get("interaction_step", 0)
logging.info(f"Resuming from step {step}, interaction step {interaction_step}")

12
lerobot/scripts/server/learner_service.py
View File

@@ -8,13 +8,13 @@ from lerobot.scripts.server.network_utils import receive_bytes_in_chunks, send_b
MAX_MESSAGE_SIZE = 4 * 1024 * 1024 # 4 MB
MAX_WORKERS = 3 # Stream parameters, send transitions and interactions
STUTDOWN_TIMEOUT = 10
SHUTDOWN_TIMEOUT = 10
class LearnerService(hilserl_pb2_grpc.LearnerServiceServicer):
def __init__(
self,
shutdown_event: Event,
shutdown_event: Event, # type: ignore
parameters_queue: Queue,
seconds_between_pushes: float,
transition_queue: Queue,
@@ -26,7 +26,7 @@ class LearnerService(hilserl_pb2_grpc.LearnerServiceServicer):
self.transition_queue = transition_queue
self.interaction_message_queue = interaction_message_queue
def StreamParameters(self, request, context):
def StreamParameters(self, request, context): # noqa: N802
# TODO: authorize the request
logging.info("[LEARNER] Received request to stream parameters from the Actor")
@@ -48,7 +48,7 @@ class LearnerService(hilserl_pb2_grpc.LearnerServiceServicer):
logging.info("[LEARNER] Stream parameters finished")
return hilserl_pb2.Empty()
def SendTransitions(self, request_iterator, _context):
def SendTransitions(self, request_iterator, _context): # noqa: N802
# TODO: authorize the request
logging.info("[LEARNER] Received request to receive transitions from the Actor")
@@ -62,7 +62,7 @@ class LearnerService(hilserl_pb2_grpc.LearnerServiceServicer):
logging.debug("[LEARNER] Finished receiving transitions")
return hilserl_pb2.Empty()
def SendInteractions(self, request_iterator, _context):
def SendInteractions(self, request_iterator, _context): # noqa: N802
# TODO: authorize the request
logging.info("[LEARNER] Received request to receive interactions from the Actor")
@@ -76,5 +76,5 @@ class LearnerService(hilserl_pb2_grpc.LearnerServiceServicer):
logging.debug("[LEARNER] Finished receiving interactions")
return hilserl_pb2.Empty()
def Ready(self, request, context):
def Ready(self, request, context): # noqa: N802
return hilserl_pb2.Empty()

47
lerobot/scripts/server/maniskill_manipulator.py
View File

@@ -219,50 +219,3 @@ def make_maniskill(
env = ManiskillMockGripperWrapper(env, nb_discrete_actions=3)
return env
# @parser.wrap()
# def main(cfg: TrainPipelineConfig):
# """Main function to run the ManiSkill environment."""
# # Create the ManiSkill environment
# env = make_maniskill(cfg.env, n_envs=1)
# # Reset the environment
# obs, info = env.reset()
# # Run a simple interaction loop
# sum_reward = 0
# for i in range(100):
# # Sample a random action
# action = env.action_space.sample()
# # Step the environment
# start_time = time.perf_counter()
# obs, reward, terminated, truncated, info = env.step(action)
# step_time = time.perf_counter() - start_time
# sum_reward += reward
# # Log information
# # Reset if episode terminated
# if terminated or truncated:
# logging.info(f"Step {i}, reward: {sum_reward}, step time: {step_time}s")
# sum_reward = 0
# obs, info = env.reset()
# # Close the environment
# env.close()
# if __name__ == "__main__":
# logging.basicConfig(level=logging.INFO)
# main()
if __name__ == "__main__":
import draccus
config = ManiskillEnvConfig()
draccus.set_config_type("json")
draccus.dump(
config=config,
stream=open(file="run_config.json", mode="w"),
)

78
lerobot/scripts/train.py
View File

@@ -53,84 +53,6 @@ from lerobot.configs.train import TrainPipelineConfig
from lerobot.scripts.eval import eval_policy
def make_optimizer_and_scheduler(cfg, policy):
if cfg.policy.name == "act":
optimizer_params_dicts = [
{
"params": [
p
for n, p in policy.named_parameters()
if not n.startswith("model.backbone") and p.requires_grad
]
},
{
"params": [
p
for n, p in policy.named_parameters()
if n.startswith("model.backbone") and p.requires_grad
],
"lr": cfg.training.lr_backbone,
},
]
optimizer = torch.optim.AdamW(
optimizer_params_dicts,
lr=cfg.training.lr,
weight_decay=cfg.training.weight_decay,
)
lr_scheduler = None
elif cfg.policy.name == "diffusion":
optimizer = torch.optim.Adam(
policy.diffusion.parameters(),
cfg.training.lr,
cfg.training.adam_betas,
cfg.training.adam_eps,
cfg.training.adam_weight_decay,
)
from diffusers.optimization import get_scheduler
lr_scheduler = get_scheduler(
cfg.training.lr_scheduler,
optimizer=optimizer,
num_warmup_steps=cfg.training.lr_warmup_steps,
num_training_steps=cfg.training.offline_steps,
)
elif policy.name == "tdmpc":
optimizer = torch.optim.Adam(policy.parameters(), cfg.training.lr)
lr_scheduler = None
elif policy.name == "sac":
optimizer = torch.optim.Adam(
[
{"params": policy.actor.parameters(), "lr": policy.config.actor_lr},
{
"params": policy.critic_ensemble.parameters(),
"lr": policy.config.critic_lr,
},
{
"params": policy.temperature.parameters(),
"lr": policy.config.temperature_lr,
},
]
)
lr_scheduler = None
elif cfg.policy.name == "vqbet":
from lerobot.common.policies.vqbet.modeling_vqbet import (
VQBeTOptimizer,
VQBeTScheduler,
)
optimizer = VQBeTOptimizer(policy, cfg)
lr_scheduler = VQBeTScheduler(optimizer, cfg)
elif cfg.policy.name == "hilserl_classifier":
optimizer = torch.optim.AdamW(policy.parameters(), cfg.policy.learning_rate)
lr_scheduler = None
else:
raise NotImplementedError()
return optimizer, lr_scheduler
def update_policy(
train_metrics: MetricsTracker,
policy: PreTrainedPolicy,

466
lerobot/scripts/train_hilserl_classifier.py
View File

@@ -1,466 +0,0 @@
# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import logging
import time
from contextlib import nullcontext
from pprint import pformat
import numpy as np
import torch
import torch.nn as nn
import wandb
from deepdiff import DeepDiff
from omegaconf import DictConfig, OmegaConf
from termcolor import colored
from torch import optim
from torch.autograd import profiler
from torch.cuda.amp import GradScaler
from torch.utils.data import DataLoader, RandomSampler, WeightedRandomSampler
from tqdm import tqdm
from lerobot.common.datasets.factory import resolve_delta_timestamps
from lerobot.common.datasets.lerobot_dataset import LeRobotDataset
from lerobot.common.policies.hilserl.classifier.configuration_classifier import ClassifierConfig
from lerobot.common.policies.hilserl.classifier.modeling_classifier import Classifier
from lerobot.common.utils.utils import (
format_big_number,
get_safe_torch_device,
init_hydra_config,
init_logging,
set_global_seed,
)
from lerobot.scripts.server.buffer import random_shift
def get_model(cfg, logger): # noqa I001
classifier_config = _policy_cfg_from_hydra_cfg(ClassifierConfig, cfg)
model = Classifier(classifier_config)
if cfg.resume:
model.load_state_dict(Classifier.from_pretrained(str(logger.last_pretrained_model_dir)).state_dict())
return model
def create_balanced_sampler(dataset, cfg):
# Get underlying dataset if using Subset
original_dataset = dataset.dataset if isinstance(dataset, torch.utils.data.Subset) else dataset
# Get indices if using Subset (for slicing)
indices = dataset.indices if isinstance(dataset, torch.utils.data.Subset) else None
# Get labels from Hugging Face dataset
if indices is not None:
# Get subset of labels using Hugging Face's select()
hf_subset = original_dataset.hf_dataset.select(indices)
labels = hf_subset[cfg.training.label_key]
else:
# Get all labels directly
labels = original_dataset.hf_dataset[cfg.training.label_key]
labels = torch.stack(labels)
_, counts = torch.unique(labels, return_counts=True)
class_weights = 1.0 / counts.float()
sample_weights = class_weights[labels]
return WeightedRandomSampler(weights=sample_weights, num_samples=len(sample_weights), replacement=True)
def support_amp(device: torch.device, cfg: DictConfig) -> bool:
# Check if the device supports AMP
# Here is an example of the issue that says that MPS doesn't support AMP properply
return cfg.training.use_amp and device.type in ("cuda", "cpu")
def train_epoch(model, train_loader, criterion, optimizer, grad_scaler, device, logger, step, cfg):
# Single epoch training loop with AMP support and progress tracking
model.train()
correct = 0
total = 0
pbar = tqdm(train_loader, desc="Training")
for batch_idx, batch in enumerate(pbar):
start_time = time.perf_counter()
images = [batch[img_key].to(device) for img_key in cfg.training.image_keys]
images = [random_shift(img, 4) for img in images]
labels = batch[cfg.training.label_key].float().to(device)
# Forward pass with optional AMP
with torch.autocast(device_type=device.type) if support_amp(device, cfg) else nullcontext():
outputs = model(images)
loss = criterion(outputs.logits, labels)
# Backward pass with gradient scaling if AMP enabled
optimizer.zero_grad()
if cfg.training.use_amp:
grad_scaler.scale(loss).backward()
grad_scaler.step(optimizer)
grad_scaler.update()
else:
loss.backward()
optimizer.step()
# Track metrics
if model.config.num_classes == 2:
predictions = (torch.sigmoid(outputs.logits) > 0.5).float()
else:
predictions = torch.argmax(outputs.logits, dim=1)
correct += (predictions == labels).sum().item()
total += labels.size(0)
current_acc = 100 * correct / total
train_info = {
"loss": loss.item(),
"accuracy": current_acc,
"dataloading_s": time.perf_counter() - start_time,
}
logger.log_dict(train_info, step + batch_idx, mode="train")
pbar.set_postfix({"loss": f"{loss.item():.4f}", "acc": f"{current_acc:.2f}%"})
def validate(model, val_loader, criterion, device, logger, cfg):
# Validation loop with metric tracking and sample logging
model.eval()
correct = 0
total = 0
batch_start_time = time.perf_counter()
samples = []
running_loss = 0
inference_times = []
with (
torch.no_grad(),
torch.autocast(device_type=device.type) if support_amp(device, cfg) else nullcontext(),
):
for batch in tqdm(val_loader, desc="Validation"):
images = [batch[img_key].to(device) for img_key in cfg.training.image_keys]
labels = batch[cfg.training.label_key].float().to(device)
if cfg.training.profile_inference_time and logger._cfg.wandb.enable:
with (
profiler.profile(record_shapes=True) as prof,
profiler.record_function("model_inference"),
):
outputs = model(images)
inference_times.append(
next(x for x in prof.key_averages() if x.key == "model_inference").cpu_time
)
else:
outputs = model(images)
loss = criterion(outputs.logits, labels)
# Track metrics
if model.config.num_classes == 2:
predictions = (torch.sigmoid(outputs.logits) > 0.5).float()
else:
predictions = torch.argmax(outputs.logits, dim=1)
correct += (predictions == labels).sum().item()
total += labels.size(0)
running_loss += loss.item()
# Log sample predictions for visualization
if len(samples) < cfg.eval.num_samples_to_log:
for i in range(min(cfg.eval.num_samples_to_log - len(samples), len(images))):
if model.config.num_classes == 2:
confidence = round(outputs.probabilities[i].item(), 3)
else:
confidence = [round(prob, 3) for prob in outputs.probabilities[i].tolist()]
samples.append(
{
**{
f"image_{img_key}": wandb.Image(images[img_idx][i].cpu())
for img_idx, img_key in enumerate(cfg.training.image_keys)
},
"true_label": labels[i].item(),
"predicted": predictions[i].item(),
"confidence": confidence,
}
)
accuracy = 100 * correct / total
avg_loss = running_loss / len(val_loader)
print(f"Average validation loss {avg_loss}, and accuracy {accuracy}")
eval_info = {
"loss": avg_loss,
"accuracy": accuracy,
"eval_s": time.perf_counter() - batch_start_time,
"eval/prediction_samples": wandb.Table(
data=[list(s.values()) for s in samples],
columns=list(samples[0].keys()),
)
if logger._cfg.wandb.enable
else None,
}
if len(inference_times) > 0:
eval_info["inference_time_avg"] = np.mean(inference_times)
eval_info["inference_time_median"] = np.median(inference_times)
eval_info["inference_time_std"] = np.std(inference_times)
eval_info["inference_time_batch_size"] = val_loader.batch_size
print(
f"Inference mean time: {eval_info['inference_time_avg']:.2f} us, median: {eval_info['inference_time_median']:.2f} us, std: {eval_info['inference_time_std']:.2f} us, with {len(inference_times)} iterations on {device.type} device, batch size: {eval_info['inference_time_batch_size']}"
)
return accuracy, eval_info
def benchmark_inference_time(model, dataset, logger, cfg, device, step):
if not cfg.training.profile_inference_time:
return
iters = cfg.training.profile_inference_time_iters
inference_times = []
loader = DataLoader(
dataset,
batch_size=1,
num_workers=cfg.training.num_workers,
sampler=RandomSampler(dataset),
pin_memory=True,
)
model.eval()
with torch.no_grad():
for _ in tqdm(range(iters), desc="Benchmarking inference time"):
x = next(iter(loader))
x = [x[img_key].to(device) for img_key in cfg.training.image_keys]
# Warm up
for _ in range(10):
_ = model(x)
# sync the device
if device.type == "cuda":
torch.cuda.synchronize()
elif device.type == "mps":
torch.mps.synchronize()
with (
profiler.profile(record_shapes=True) as prof,
profiler.record_function("model_inference"),
):
_ = model(x)
inference_times.append(
next(x for x in prof.key_averages() if x.key == "model_inference").cpu_time
)
inference_times = np.array(inference_times)
avg, median, std = (
inference_times.mean(),
np.median(inference_times),
inference_times.std(),
)
print(
f"Inference time mean: {avg:.2f} us, median: {median:.2f} us, std: {std:.2f} us, with {iters} iterations on {device.type} device"
)
if logger._cfg.wandb.enable:
logger.log_dict(
{
"inference_time_benchmark_avg": avg,
"inference_time_benchmark_median": median,
"inference_time_benchmark_std": std,
},
step + 1,
mode="eval",
)
return avg, median, std
def train(cfg: DictConfig, out_dir: str | None = None, job_name: str | None = None) -> None:
if out_dir is None:
raise NotImplementedError()
if job_name is None:
raise NotImplementedError()
# Main training pipeline with support for resuming training
init_logging()
logging.info(OmegaConf.to_yaml(cfg))
# Initialize training environment
device = get_safe_torch_device(cfg.device, log=True)
set_global_seed(cfg.seed)
# Setup dataset and dataloaders
dataset = LeRobotDataset(
cfg.dataset_repo_id,
root=cfg.dataset_root,
local_files_only=cfg.local_files_only,
)
logging.info(f"Dataset size: {len(dataset)}")
n_total = len(dataset)
n_train = int(cfg.train_split_proportion * len(dataset))
train_dataset = torch.utils.data.Subset(dataset, range(0, n_train))
val_dataset = torch.utils.data.Subset(dataset, range(n_train, n_total))
sampler = create_balanced_sampler(train_dataset, cfg)
train_loader = DataLoader(
train_dataset,
batch_size=cfg.training.batch_size,
num_workers=cfg.training.num_workers,
sampler=sampler,
pin_memory=device.type == "cuda",
)
val_loader = DataLoader(
val_dataset,
batch_size=cfg.eval.batch_size,
shuffle=False,
num_workers=cfg.training.num_workers,
pin_memory=device.type == "cuda",
)
# Resume training if requested
step = 0
best_val_acc = 0
if cfg.resume:
if not Logger.get_last_checkpoint_dir(out_dir).exists():
raise RuntimeError(
"You have set resume=True, but there is no model checkpoint in "
f"{Logger.get_last_checkpoint_dir(out_dir)}"
)
checkpoint_cfg_path = str(Logger.get_last_pretrained_model_dir(out_dir) / "config.yaml")
logging.info(
colored(
"You have set resume=True, indicating that you wish to resume a run",
color="yellow",
attrs=["bold"],
)
)
# Load and validate checkpoint configuration
checkpoint_cfg = init_hydra_config(checkpoint_cfg_path)
# Check for differences between the checkpoint configuration and provided configuration.
# Hack to resolve the delta_timestamps ahead of time in order to properly diff.
resolve_delta_timestamps(cfg)
diff = DeepDiff(OmegaConf.to_container(checkpoint_cfg), OmegaConf.to_container(cfg))
# Ignore the `resume` and parameters.
if "values_changed" in diff and "root['resume']" in diff["values_changed"]:
del diff["values_changed"]["root['resume']"]
if len(diff) > 0:
logging.warning(
"At least one difference was detected between the checkpoint configuration and "
f"the provided configuration: \n{pformat(diff)}\nNote that the checkpoint configuration "
"takes precedence.",
)
# Use the checkpoint config instead of the provided config (but keep `resume` parameter).
cfg = checkpoint_cfg
cfg.resume = True
# Initialize model and training components
model = get_model(cfg=cfg, logger=logger).to(device)
optimizer = optim.AdamW(model.parameters(), lr=cfg.training.learning_rate)
# Use BCEWithLogitsLoss for binary classification and CrossEntropyLoss for multi-class
criterion = nn.BCEWithLogitsLoss() if model.config.num_classes == 2 else nn.CrossEntropyLoss()
grad_scaler = GradScaler(enabled=cfg.training.use_amp)
# Log model parameters
num_learnable_params = sum(p.numel() for p in model.parameters() if p.requires_grad)
num_total_params = sum(p.numel() for p in model.parameters())
logging.info(f"Learnable parameters: {format_big_number(num_learnable_params)}")
logging.info(f"Total parameters: {format_big_number(num_total_params)}")
if cfg.resume:
step = logger.load_last_training_state(optimizer, None)
# Training loop with validation and checkpointing
for epoch in range(cfg.training.num_epochs):
logging.info(f"\nEpoch {epoch + 1}/{cfg.training.num_epochs}")
train_epoch(
model,
train_loader,
criterion,
optimizer,
grad_scaler,
device,
logger,
step,
cfg,
)
# Periodic validation
if cfg.training.eval_freq > 0 and (epoch + 1) % cfg.training.eval_freq == 0:
val_acc, eval_info = validate(
model,
val_loader,
criterion,
device,
logger,
cfg,
)
logger.log_dict(eval_info, step + len(train_loader), mode="eval")
# Save best model
if val_acc > best_val_acc:
best_val_acc = val_acc
logger.save_checkpoint(
train_step=step + len(train_loader),
policy=model,
optimizer=optimizer,
scheduler=None,
identifier="best",
)
# Periodic checkpointing
if cfg.training.save_checkpoint and (epoch + 1) % cfg.training.save_freq == 0:
logger.save_checkpoint(
train_step=step + len(train_loader),
policy=model,
optimizer=optimizer,
scheduler=None,
identifier=f"{epoch + 1:06d}",
)
step += len(train_loader)
benchmark_inference_time(model, dataset, logger, cfg, device, step)
logging.info("Training completed")
@hydra.main(
version_base="1.2",
config_name="hilserl_classifier",
config_path="../configs/policy",
)
def train_cli(cfg: dict):
train(
cfg,
out_dir=hydra.core.hydra_config.HydraConfig.get().run.dir,
job_name=hydra.core.hydra_config.HydraConfig.get().job.name,
)
def train_notebook(
out_dir=None,
job_name=None,
config_name="hilserl_classifier",
config_path="../configs/policy",
):
from hydra import compose, initialize
hydra.core.global_hydra.GlobalHydra.instance().clear()
initialize(config_path=config_path)
cfg = compose(config_name=config_name)
train(cfg, out_dir=out_dir, job_name=job_name)
if __name__ == "__main__":
train_cli()

594
lerobot/scripts/train_sac.py
View File

@@ -1,594 +0,0 @@
#!/usr/bin/env python
# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import functools
import logging
import random
from pprint import pformat
from typing import Callable, Optional, Sequence, TypedDict
import hydra
import torch
import torch.nn.functional as F
from omegaconf import DictConfig, OmegaConf
from torch import nn
from tqdm import tqdm
# TODO: Remove the import of maniskill
from lerobot.common.datasets.factory import make_dataset
from lerobot.common.datasets.lerobot_dataset import LeRobotDataset
from lerobot.common.envs.factory import make_maniskill_env
from lerobot.common.envs.utils import preprocess_maniskill_observation
from lerobot.common.logger import Logger, log_output_dir
from lerobot.common.policies.factory import make_policy
from lerobot.common.policies.sac.modeling_sac import SACPolicy
from lerobot.common.utils.utils import (
format_big_number,
get_safe_torch_device,
init_logging,
set_global_seed,
)
def make_optimizers_and_scheduler(cfg, policy):
optimizer_actor = torch.optim.Adam(
# NOTE: Handle the case of shared encoder where the encoder weights are not optimized with the gradient of the actor
params=policy.actor.parameters_to_optimize,
lr=policy.config.actor_lr,
)
optimizer_critic = torch.optim.Adam(
params=policy.critic_ensemble.parameters(), lr=policy.config.critic_lr
)
# We wrap policy log temperature in list because this is a torch tensor and not a nn.Module
optimizer_temperature = torch.optim.Adam(params=[policy.log_alpha], lr=policy.config.critic_lr)
lr_scheduler = None
optimizers = {
"actor": optimizer_actor,
"critic": optimizer_critic,
"temperature": optimizer_temperature,
}
return optimizers, lr_scheduler
class Transition(TypedDict):
state: dict[str, torch.Tensor]
action: torch.Tensor
reward: float
next_state: dict[str, torch.Tensor]
done: bool
complementary_info: dict[str, torch.Tensor] = None
class BatchTransition(TypedDict):
state: dict[str, torch.Tensor]
action: torch.Tensor
reward: torch.Tensor
next_state: dict[str, torch.Tensor]
done: torch.Tensor
def random_crop_vectorized(images: torch.Tensor, output_size: tuple) -> torch.Tensor:
"""
Perform a per-image random crop over a batch of images in a vectorized way.
(Same as shown previously.)
"""
B, C, H, W = images.shape
crop_h, crop_w = output_size
if crop_h > H or crop_w > W:
raise ValueError(
f"Requested crop size ({crop_h}, {crop_w}) is bigger than the image size ({H}, {W})."
)
tops = torch.randint(0, H - crop_h + 1, (B,), device=images.device)
lefts = torch.randint(0, W - crop_w + 1, (B,), device=images.device)
rows = torch.arange(crop_h, device=images.device).unsqueeze(0) + tops.unsqueeze(1)
cols = torch.arange(crop_w, device=images.device).unsqueeze(0) + lefts.unsqueeze(1)
rows = rows.unsqueeze(2).expand(-1, -1, crop_w) # (B, crop_h, crop_w)
cols = cols.unsqueeze(1).expand(-1, crop_h, -1) # (B, crop_h, crop_w)
images_hwcn = images.permute(0, 2, 3, 1) # (B, H, W, C)
# Gather pixels
cropped_hwcn = images_hwcn[torch.arange(B, device=images.device).view(B, 1, 1), rows, cols, :]
# cropped_hwcn => (B, crop_h, crop_w, C)
cropped = cropped_hwcn.permute(0, 3, 1, 2) # (B, C, crop_h, crop_w)
return cropped
def random_shift(images: torch.Tensor, pad: int = 4):
"""Vectorized random shift, imgs: (B,C,H,W), pad: #pixels"""
_, _, h, w = images.shape
images = F.pad(input=images, pad=(pad, pad, pad, pad), mode="replicate")
return random_crop_vectorized(images=images, output_size=(h, w))
class ReplayBuffer:
def __init__(
self,
capacity: int,
device: str = "cuda:0",
state_keys: Optional[Sequence[str]] = None,
image_augmentation_function: Optional[Callable] = None,
use_drq: bool = True,
):
"""
Args:
capacity (int): Maximum number of transitions to store in the buffer.
device (str): The device where the tensors will be moved ("cuda:0" or "cpu").
state_keys (List[str]): The list of keys that appear in `state` and `next_state`.
image_augmentation_function (Optional[Callable]): A function that takes a batch of images
and returns a batch of augmented images. If None, a default augmentation function is used.
use_drq (bool): Whether to use the default DRQ image augmentation style, when sampling in the buffer.
"""
self.capacity = capacity
self.device = device
self.memory: list[Transition] = []
self.position = 0
# If no state_keys provided, default to an empty list
# (you can handle this differently if needed)
self.state_keys = state_keys if state_keys is not None else []
if image_augmentation_function is None:
self.image_augmentation_function = functools.partial(random_shift, pad=4)
self.use_drq = use_drq
def add(
self,
state: dict[str, torch.Tensor],
action: torch.Tensor,
reward: float,
next_state: dict[str, torch.Tensor],
done: bool,
complementary_info: Optional[dict[str, torch.Tensor]] = None,
):
"""Saves a transition."""
if len(self.memory) < self.capacity:
self.memory.append(None)
# Create and store the Transition
self.memory[self.position] = Transition(
state=state,
action=action,
reward=reward,
next_state=next_state,
done=done,
complementary_info=complementary_info,
)
self.position: int = (self.position + 1) % self.capacity
# TODO: ADD image_augmentation and use_drq arguments in this function in order to instantiate the class with them
@classmethod
def from_lerobot_dataset(
cls,
lerobot_dataset: LeRobotDataset,
device: str = "cuda:0",
state_keys: Optional[Sequence[str]] = None,
) -> "ReplayBuffer":
"""
Convert a LeRobotDataset into a ReplayBuffer.
Args:
lerobot_dataset (LeRobotDataset): The dataset to convert.
device (str): The device . Defaults to "cuda:0".
state_keys (Optional[Sequence[str]], optional): The list of keys that appear in `state` and `next_state`.
Defaults to None.
Returns:
ReplayBuffer: The replay buffer with offline dataset transitions.
"""
# We convert the LeRobotDataset into a replay buffer, because it is more efficient to sample from
# a replay buffer than from a lerobot dataset.
replay_buffer = cls(capacity=len(lerobot_dataset), device=device, state_keys=state_keys)
list_transition = cls._lerobotdataset_to_transitions(dataset=lerobot_dataset, state_keys=state_keys)
# Fill the replay buffer with the lerobot dataset transitions
for data in list_transition:
replay_buffer.add(
state=data["state"],
action=data["action"],
reward=data["reward"],
next_state=data["next_state"],
done=data["done"],
)
return replay_buffer
@staticmethod
def _lerobotdataset_to_transitions(
dataset: LeRobotDataset,
state_keys: Optional[Sequence[str]] = None,
) -> list[Transition]:
"""
Convert a LeRobotDataset into a list of RL (s, a, r, s', done) transitions.
Args:
dataset (LeRobotDataset):
The dataset to convert. Each item in the dataset is expected to have
at least the following keys:
{
"action": ...
"next.reward": ...
"next.done": ...
"episode_index": ...
}
plus whatever your 'state_keys' specify.
state_keys (Optional[Sequence[str]]):
The dataset keys to include in 'state' and 'next_state'. Their names
will be kept as-is in the output transitions. E.g.
["observation.state", "observation.environment_state"].
If None, you must handle or define default keys.
Returns:
transitions (List[Transition]):
A list of Transition dictionaries with the same length as `dataset`.
"""
# If not provided, you can either raise an error or define a default:
if state_keys is None:
raise ValueError("You must provide a list of keys in `state_keys` that define your 'state'.")
transitions: list[Transition] = []
num_frames = len(dataset)
for i in tqdm(range(num_frames)):
current_sample = dataset[i]
# ----- 1) Current state -----
current_state: dict[str, torch.Tensor] = {}
for key in state_keys:
val = current_sample[key]
current_state[key] = val.unsqueeze(0) # Add batch dimension
# ----- 2) Action -----
action = current_sample["action"].unsqueeze(0) # Add batch dimension
# ----- 3) Reward and done -----
reward = float(current_sample["next.reward"].item()) # ensure float
done = bool(current_sample["next.done"].item()) # ensure bool
# ----- 4) Next state -----
# If not done and the next sample is in the same episode, we pull the next sample's state.
# Otherwise (done=True or next sample crosses to a new episode), next_state = current_state.
next_state = current_state # default
if not done and (i < num_frames - 1):
next_sample = dataset[i + 1]
if next_sample["episode_index"] == current_sample["episode_index"]:
# Build next_state from the same keys
next_state_data: dict[str, torch.Tensor] = {}
for key in state_keys:
val = next_sample[key]
next_state_data[key] = val.unsqueeze(0) # Add batch dimension
next_state = next_state_data
# ----- Construct the Transition -----
transition = Transition(
state=current_state,
action=action,
reward=reward,
next_state=next_state,
done=done,
)
transitions.append(transition)
return transitions
def sample(self, batch_size: int) -> BatchTransition:
"""Sample a random batch of transitions and collate them into batched tensors."""
list_of_transitions = random.sample(self.memory, batch_size)
# -- Build batched states --
batch_state = {}
for key in self.state_keys:
batch_state[key] = torch.cat([t["state"][key] for t in list_of_transitions], dim=0).to(
self.device
)
if key.startswith("observation.image") and self.use_drq:
batch_state[key] = self.image_augmentation_function(batch_state[key])
# -- Build batched actions --
batch_actions = torch.cat([t["action"] for t in list_of_transitions]).to(self.device)
# -- Build batched rewards --
batch_rewards = torch.tensor([t["reward"] for t in list_of_transitions], dtype=torch.float32).to(
self.device
)
# -- Build batched next states --
batch_next_state = {}
for key in self.state_keys:
batch_next_state[key] = torch.cat([t["next_state"][key] for t in list_of_transitions], dim=0).to(
self.device
)
if key.startswith("observation.image") and self.use_drq:
batch_next_state[key] = self.image_augmentation_function(batch_next_state[key])
# -- Build batched dones --
batch_dones = torch.tensor([t["done"] for t in list_of_transitions], dtype=torch.float32).to(
self.device
)
batch_dones = torch.tensor([t["done"] for t in list_of_transitions], dtype=torch.float32).to(
self.device
)
# Return a BatchTransition typed dict
return BatchTransition(
state=batch_state,
action=batch_actions,
reward=batch_rewards,
next_state=batch_next_state,
done=batch_dones,
)
def concatenate_batch_transitions(
left_batch_transitions: BatchTransition, right_batch_transition: BatchTransition
) -> BatchTransition:
"""NOTE: Be careful it change the left_batch_transitions in place"""
left_batch_transitions["state"] = {
key: torch.cat(
[
left_batch_transitions["state"][key],
right_batch_transition["state"][key],
],
dim=0,
)
for key in left_batch_transitions["state"]
}
left_batch_transitions["action"] = torch.cat(
[left_batch_transitions["action"], right_batch_transition["action"]], dim=0
)
left_batch_transitions["reward"] = torch.cat(
[left_batch_transitions["reward"], right_batch_transition["reward"]], dim=0
)
left_batch_transitions["next_state"] = {
key: torch.cat(
[
left_batch_transitions["next_state"][key],
right_batch_transition["next_state"][key],
],
dim=0,
)
for key in left_batch_transitions["next_state"]
}
left_batch_transitions["done"] = torch.cat(
[left_batch_transitions["done"], right_batch_transition["done"]], dim=0
)
return left_batch_transitions
def train(cfg: DictConfig, out_dir: str | None = None, job_name: str | None = None):
if out_dir is None:
raise NotImplementedError()
if job_name is None:
raise NotImplementedError()
init_logging()
logging.info(pformat(OmegaConf.to_container(cfg)))
# Create an env dedicated to online episodes collection from policy rollout.
# online_env = make_env(cfg, n_envs=cfg.training.online_rollout_batch_size)
# NOTE: Off policy algorithm are efficient enought to use a single environment
logging.info("make_env online")
# online_env = make_env(cfg, n_envs=1)
# TODO: Remove the import of maniskill and unifiy with make env
online_env = make_maniskill_env(cfg, n_envs=1)
if cfg.training.eval_freq > 0:
logging.info("make_env eval")
# eval_env = make_env(cfg, n_envs=1)
# TODO: Remove the import of maniskill and unifiy with make env
eval_env = make_maniskill_env(cfg, n_envs=1)
# TODO: Add a way to resume training
# log metrics to terminal and wandb
logger = Logger(cfg, out_dir, wandb_job_name=job_name)
set_global_seed(cfg.seed)
# Check device is available
device = get_safe_torch_device(cfg.device, log=True)
torch.backends.cudnn.benchmark = True
torch.backends.cuda.matmul.allow_tf32 = True
logging.info("make_policy")
# TODO: At some point we should just need make sac policy
policy: SACPolicy = make_policy(
hydra_cfg=cfg,
# dataset_stats=offline_dataset.meta.stats if not cfg.resume else None,
# Hack: But if we do online traning, we do not need dataset_stats
dataset_stats=None,
pretrained_policy_name_or_path=str(logger.last_pretrained_model_dir) if cfg.resume else None,
device=device,
)
assert isinstance(policy, nn.Module)
optimizers, lr_scheduler = make_optimizers_and_scheduler(cfg, policy)
# TODO: Handle resume
num_learnable_params = sum(p.numel() for p in policy.parameters() if p.requires_grad)
num_total_params = sum(p.numel() for p in policy.parameters())
log_output_dir(out_dir)
logging.info(f"{cfg.env.task=}")
logging.info(f"{cfg.training.online_steps=}")
logging.info(f"{num_learnable_params=} ({format_big_number(num_learnable_params)})")
logging.info(f"{num_total_params=} ({format_big_number(num_total_params)})")
obs, info = online_env.reset()
# HACK for maniskill
# obs = preprocess_observation(obs)
obs = preprocess_maniskill_observation(obs)
obs = {key: obs[key].to(device, non_blocking=True) for key in obs}
replay_buffer = ReplayBuffer(
capacity=cfg.training.online_buffer_capacity,
device=device,
state_keys=cfg.policy.input_shapes.keys(),
)
batch_size = cfg.training.batch_size
if cfg.dataset_repo_id is not None:
logging.info("make_dataset offline buffer")
offline_dataset = make_dataset(cfg)
logging.info("Convertion to a offline replay buffer")
offline_replay_buffer = ReplayBuffer.from_lerobot_dataset(
offline_dataset, device=device, state_keys=cfg.policy.input_shapes.keys()
)
batch_size: int = batch_size // 2 # We will sample from both replay buffer
# NOTE: For the moment we will solely handle the case of a single environment
sum_reward_episode = 0
for interaction_step in range(cfg.training.online_steps):
# NOTE: At some point we should use a wrapper to handle the observation
if interaction_step >= cfg.training.online_step_before_learning:
action = policy.select_action(batch=obs)
next_obs, reward, done, truncated, info = online_env.step(action.cpu().numpy())
else:
action = online_env.action_space.sample()
next_obs, reward, done, truncated, info = online_env.step(action)
# HACK
action = torch.tensor(action, dtype=torch.float32).to(device, non_blocking=True)
# HACK: For maniskill
# next_obs = preprocess_observation(next_obs)
next_obs = preprocess_maniskill_observation(next_obs)
next_obs = {key: next_obs[key].to(device, non_blocking=True) for key in obs}
sum_reward_episode += float(reward[0])
# Because we are using a single environment
# we can safely assume that the episode is done
if done[0] or truncated[0]:
logging.info(f"Global step {interaction_step}: Episode reward: {sum_reward_episode}")
logger.log_dict({"Sum episode reward": sum_reward_episode}, interaction_step)
sum_reward_episode = 0
# HACK: This is for maniskill
logging.info(
f"global step {interaction_step}: episode success: {info['success'].float().item()} \n"
)
logger.log_dict({"Episode success": info["success"].float().item()}, interaction_step)
replay_buffer.add(
state=obs,
action=action,
reward=float(reward[0]),
next_state=next_obs,
done=done[0],
)
obs = next_obs
if interaction_step < cfg.training.online_step_before_learning:
continue
for _ in range(cfg.policy.utd_ratio - 1):
batch = replay_buffer.sample(batch_size)
if cfg.dataset_repo_id is not None:
batch_offline = offline_replay_buffer.sample(batch_size)
batch = concatenate_batch_transitions(batch, batch_offline)
actions = batch["action"]
rewards = batch["reward"]
observations = batch["state"]
next_observations = batch["next_state"]
done = batch["done"]
loss_critic = policy.compute_loss_critic(
observations=observations,
actions=actions,
rewards=rewards,
next_observations=next_observations,
done=done,
)
optimizers["critic"].zero_grad()
loss_critic.backward()
optimizers["critic"].step()
batch = replay_buffer.sample(batch_size)
if cfg.dataset_repo_id is not None:
batch_offline = offline_replay_buffer.sample(batch_size)
batch = concatenate_batch_transitions(
left_batch_transitions=batch, right_batch_transition=batch_offline
)
actions = batch["action"]
rewards = batch["reward"]
observations = batch["state"]
next_observations = batch["next_state"]
done = batch["done"]
loss_critic = policy.compute_loss_critic(
observations=observations,
actions=actions,
rewards=rewards,
next_observations=next_observations,
done=done,
)
optimizers["critic"].zero_grad()
loss_critic.backward()
optimizers["critic"].step()
training_infos = {}
training_infos["loss_critic"] = loss_critic.item()
if interaction_step % cfg.training.policy_update_freq == 0:
# TD3 Trick
for _ in range(cfg.training.policy_update_freq):
loss_actor = policy.compute_loss_actor(observations=observations)
optimizers["actor"].zero_grad()
loss_actor.backward()
optimizers["actor"].step()
training_infos["loss_actor"] = loss_actor.item()
loss_temperature = policy.compute_loss_temperature(observations=observations)
optimizers["temperature"].zero_grad()
loss_temperature.backward()
optimizers["temperature"].step()
training_infos["loss_temperature"] = loss_temperature.item()
if interaction_step % cfg.training.log_freq == 0:
logger.log_dict(training_infos, interaction_step, mode="train")
policy.update_target_networks()
@hydra.main(version_base="1.2", config_name="default", config_path="../configs")
def train_cli(cfg: dict):
train(
cfg,
out_dir=hydra.core.hydra_config.HydraConfig.get().run.dir,
job_name=hydra.core.hydra_config.HydraConfig.get().job.name,
)
def train_notebook(out_dir=None, job_name=None, config_name="default", config_path="../configs"):
from hydra import compose, initialize
hydra.core.global_hydra.GlobalHydra.instance().clear()
initialize(config_path=config_path)
cfg = compose(config_name=config_name)
train(cfg, out_dir=out_dir, job_name=job_name)
if __name__ == "__main__":
train_cli()

4
tests/optim/test_optimizers.py
View File

@@ -168,7 +168,7 @@ def test_save_multi_optimizer_state(multi_optimizers, tmp_path):
save_optimizer_state(multi_optimizers, tmp_path)
# Verify that directories were created for each optimizer
for name in multi_optimizers.keys():
for name in multi_optimizers:
assert (tmp_path / name).is_dir()
assert (tmp_path / name / OPTIMIZER_STATE).is_file()
assert (tmp_path / name / OPTIMIZER_PARAM_GROUPS).is_file()
@@ -204,7 +204,7 @@ def test_save_and_load_multi_optimizer_state(base_params_dict, multi_optimizers,
loaded_optimizers = load_optimizer_state(new_optimizers, tmp_path)
# Verify state dictionaries match
for name in multi_optimizers.keys():
for name in multi_optimizers:
torch.testing.assert_close(multi_optimizers[name].state_dict(), loaded_optimizers[name].state_dict())