Update configs to match new LeRobot RL refactor
#3
by lilkm HF Staff - opened
- reward_classifier/config.json +43 -51
- rl/env_config.json +37 -45
- rl/gym_hil/env_config.json +56 -9
- rl/gym_hil/train_config.json +137 -80
- rl/train_config.json +103 -155
reward_classifier/config.json
CHANGED
|
@@ -1,54 +1,46 @@
|
|
| 1 |
{
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
},
|
| 6 |
-
"
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
"observation.images.front": {
|
| 27 |
-
"type": "VISUAL",
|
| 28 |
-
"shape": [
|
| 29 |
-
3,
|
| 30 |
-
128,
|
| 31 |
-
128
|
| 32 |
-
]
|
| 33 |
-
}
|
| 34 |
-
}
|
| 35 |
-
},
|
| 36 |
-
"batch_size": 16,
|
| 37 |
-
"num_workers": 4,
|
| 38 |
-
"steps": 5000,
|
| 39 |
-
"log_freq": 10,
|
| 40 |
-
"eval_freq": 1000,
|
| 41 |
-
"save_freq": 1000,
|
| 42 |
-
"save_checkpoint": true,
|
| 43 |
-
"seed": 2,
|
| 44 |
-
"resume": false,
|
| 45 |
-
"optimizer": {
|
| 46 |
-
"grad_clip_norm": 10.0
|
| 47 |
-
},
|
| 48 |
-
"wandb": {
|
| 49 |
-
"enable": true,
|
| 50 |
-
"project": "reward-classifier",
|
| 51 |
-
"disable_artifact": false
|
| 52 |
-
},
|
| 53 |
-
"job_name": "reward-classifier"
|
| 54 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"dataset": {
|
| 3 |
+
"repo_id": "lilkm/pick_cube_franka_panda_30",
|
| 4 |
+
"root": null
|
| 5 |
+
},
|
| 6 |
+
"policy": {
|
| 7 |
+
"type": "reward_classifier",
|
| 8 |
+
"model_name": "lerobot/resnet10",
|
| 9 |
+
"model_type": "cnn",
|
| 10 |
+
"num_cameras": 2,
|
| 11 |
+
"num_classes": 2,
|
| 12 |
+
"hidden_dim": 256,
|
| 13 |
+
"dropout_rate": 0.1,
|
| 14 |
+
"learning_rate": 1e-4,
|
| 15 |
+
"device": "cuda",
|
| 16 |
+
"input_features": {
|
| 17 |
+
"observation.images.front": {
|
| 18 |
+
"type": "VISUAL",
|
| 19 |
+
"shape": [3, 128, 128]
|
| 20 |
+
},
|
| 21 |
+
"observation.images.wrist": {
|
| 22 |
+
"type": "VISUAL",
|
| 23 |
+
"shape": [3, 128, 128]
|
| 24 |
+
}
|
| 25 |
},
|
| 26 |
+
"push_to_hub": false,
|
| 27 |
+
"repo_id": "lilkm/pick_cube_reward_classifier"
|
| 28 |
+
},
|
| 29 |
+
"batch_size": 16,
|
| 30 |
+
"num_workers": 4,
|
| 31 |
+
"steps": 5000,
|
| 32 |
+
"log_freq": 10,
|
| 33 |
+
"eval_freq": 1000,
|
| 34 |
+
"save_freq": 1000,
|
| 35 |
+
"save_checkpoint": true,
|
| 36 |
+
"seed": 2,
|
| 37 |
+
"resume": false,
|
| 38 |
+
"optimizer": {
|
| 39 |
+
"grad_clip_norm": 10.0
|
| 40 |
+
},
|
| 41 |
+
"wandb": {
|
| 42 |
+
"enable": false,
|
| 43 |
+
"project": "reward-classifier"
|
| 44 |
+
},
|
| 45 |
+
"job_name": "reward-classifier"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
}
|
rl/env_config.json
CHANGED
|
@@ -1,16 +1,28 @@
|
|
| 1 |
{
|
| 2 |
"env": {
|
|
|
|
|
|
|
| 3 |
"robot": {
|
| 4 |
"type": "so100_follower",
|
| 5 |
-
"port": "/dev/
|
| 6 |
-
"
|
|
|
|
| 7 |
"cameras": {
|
| 8 |
"top": {
|
| 9 |
"type": "opencv",
|
| 10 |
-
"index_or_path":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
"height": 480,
|
| 12 |
"width": 640,
|
| 13 |
-
"fps": 30
|
|
|
|
| 14 |
}
|
| 15 |
}
|
| 16 |
},
|
|
@@ -20,75 +32,55 @@
|
|
| 20 |
},
|
| 21 |
"processor": {
|
| 22 |
"control_mode": "gamepad",
|
|
|
|
| 23 |
"observation": {
|
| 24 |
-
"display_cameras":
|
| 25 |
"add_joint_velocity_to_observation": true,
|
| 26 |
"add_current_to_observation": true,
|
| 27 |
-
"add_ee_pose_to_observation":
|
| 28 |
},
|
| 29 |
"image_preprocessing": {
|
| 30 |
-
"crop_params_dict":
|
| 31 |
-
|
| 32 |
-
},
|
| 33 |
-
"resize_size": [128, 128]
|
| 34 |
},
|
| 35 |
"gripper": {
|
| 36 |
"use_gripper": true,
|
| 37 |
-
"gripper_penalty": -0.02
|
| 38 |
-
"gripper_penalty_in_reward": false
|
| 39 |
},
|
| 40 |
"reset": {
|
| 41 |
-
"fixed_reset_joint_positions": [
|
| 42 |
-
0.0,
|
| 43 |
-
0.0,
|
| 44 |
-
0.0,
|
| 45 |
-
90.0,
|
| 46 |
-
0.0,
|
| 47 |
-
30.0
|
| 48 |
-
],
|
| 49 |
"reset_time_s": 2.5,
|
| 50 |
-
"control_time_s":
|
| 51 |
"terminate_on_success": true
|
| 52 |
},
|
| 53 |
"inverse_kinematics": {
|
| 54 |
-
"urdf_path": "/
|
| 55 |
"target_frame_name": "gripper_frame_link",
|
| 56 |
"end_effector_bounds": {
|
| 57 |
-
"min": [
|
| 58 |
-
|
| 59 |
-
-0.065,
|
| 60 |
-
0.005
|
| 61 |
-
],
|
| 62 |
-
"max": [
|
| 63 |
-
0.25,
|
| 64 |
-
0.065,
|
| 65 |
-
0.1
|
| 66 |
-
]
|
| 67 |
},
|
| 68 |
"end_effector_step_sizes": {
|
| 69 |
-
"x": 0.
|
| 70 |
-
"y": 0.
|
| 71 |
-
"z": 0.
|
| 72 |
}
|
| 73 |
},
|
| 74 |
"reward_classifier": {
|
| 75 |
"pretrained_path": null,
|
| 76 |
"success_threshold": 0.5,
|
| 77 |
"success_reward": 1.0
|
| 78 |
-
}
|
| 79 |
-
|
| 80 |
-
},
|
| 81 |
-
"name": "real_robot",
|
| 82 |
-
"fps": 10
|
| 83 |
},
|
| 84 |
"dataset": {
|
| 85 |
-
"repo_id": "
|
| 86 |
"root": null,
|
| 87 |
-
"task": "",
|
| 88 |
-
"num_episodes_to_record":
|
| 89 |
"replay_episode": null,
|
| 90 |
"push_to_hub": true
|
| 91 |
},
|
| 92 |
-
"mode":
|
| 93 |
-
"device": "
|
| 94 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"env": {
|
| 3 |
+
"name": "real_robot",
|
| 4 |
+
"fps": 10,
|
| 5 |
"robot": {
|
| 6 |
"type": "so100_follower",
|
| 7 |
+
"port": "/dev/ttyACM0",
|
| 8 |
+
"use_degrees": true,
|
| 9 |
+
"id": "blue_follower",
|
| 10 |
"cameras": {
|
| 11 |
"top": {
|
| 12 |
"type": "opencv",
|
| 13 |
+
"index_or_path": "/dev/video2",
|
| 14 |
+
"height": 480,
|
| 15 |
+
"width": 640,
|
| 16 |
+
"fps": 30,
|
| 17 |
+
"fourcc": "MJPG"
|
| 18 |
+
},
|
| 19 |
+
"wrist": {
|
| 20 |
+
"type": "opencv",
|
| 21 |
+
"index_or_path": "/dev/video4",
|
| 22 |
"height": 480,
|
| 23 |
"width": 640,
|
| 24 |
+
"fps": 30,
|
| 25 |
+
"fourcc": "MJPG"
|
| 26 |
}
|
| 27 |
}
|
| 28 |
},
|
|
|
|
| 32 |
},
|
| 33 |
"processor": {
|
| 34 |
"control_mode": "gamepad",
|
| 35 |
+
"max_gripper_pos": 30,
|
| 36 |
"observation": {
|
| 37 |
+
"display_cameras": true,
|
| 38 |
"add_joint_velocity_to_observation": true,
|
| 39 |
"add_current_to_observation": true,
|
| 40 |
+
"add_ee_pose_to_observation": false
|
| 41 |
},
|
| 42 |
"image_preprocessing": {
|
| 43 |
+
"crop_params_dict": null,
|
| 44 |
+
"resize_size": [128, 128]
|
|
|
|
|
|
|
| 45 |
},
|
| 46 |
"gripper": {
|
| 47 |
"use_gripper": true,
|
| 48 |
+
"gripper_penalty": -0.02
|
|
|
|
| 49 |
},
|
| 50 |
"reset": {
|
| 51 |
+
"fixed_reset_joint_positions": [0.0, -20.0, 25.0, 90.0, 90.0, 30.0],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 52 |
"reset_time_s": 2.5,
|
| 53 |
+
"control_time_s": 30.0,
|
| 54 |
"terminate_on_success": true
|
| 55 |
},
|
| 56 |
"inverse_kinematics": {
|
| 57 |
+
"urdf_path": "./SO101/so101_new_calib.urdf",
|
| 58 |
"target_frame_name": "gripper_frame_link",
|
| 59 |
"end_effector_bounds": {
|
| 60 |
+
"min": [0.115, -0.145, -0.0075],
|
| 61 |
+
"max": [0.26, 0.145, 0.1]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
},
|
| 63 |
"end_effector_step_sizes": {
|
| 64 |
+
"x": 0.009,
|
| 65 |
+
"y": 0.009,
|
| 66 |
+
"z": 0.01
|
| 67 |
}
|
| 68 |
},
|
| 69 |
"reward_classifier": {
|
| 70 |
"pretrained_path": null,
|
| 71 |
"success_threshold": 0.5,
|
| 72 |
"success_reward": 1.0
|
| 73 |
+
}
|
| 74 |
+
}
|
|
|
|
|
|
|
|
|
|
| 75 |
},
|
| 76 |
"dataset": {
|
| 77 |
+
"repo_id": "lilkm/so100_gamepad_insert_high_quality",
|
| 78 |
"root": null,
|
| 79 |
+
"task": "insert usb",
|
| 80 |
+
"num_episodes_to_record": 30,
|
| 81 |
"replay_episode": null,
|
| 82 |
"push_to_hub": true
|
| 83 |
},
|
| 84 |
+
"mode": "record",
|
| 85 |
+
"device": "cuda"
|
| 86 |
}
|
rl/gym_hil/env_config.json
CHANGED
|
@@ -7,27 +7,74 @@
|
|
| 7 |
"teleop": null,
|
| 8 |
"processor": {
|
| 9 |
"control_mode": "gamepad",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
"gripper": {
|
| 11 |
"use_gripper": true,
|
| 12 |
-
"gripper_penalty": -0.02
|
| 13 |
-
"gripper_penalty_in_reward": false
|
| 14 |
},
|
| 15 |
"reset": {
|
| 16 |
-
"fixed_reset_joint_positions": [
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
}
|
| 22 |
},
|
| 23 |
"dataset": {
|
| 24 |
-
"repo_id": "
|
| 25 |
"root": null,
|
| 26 |
"task": "PandaPickCubeGamepad-v0",
|
| 27 |
"num_episodes_to_record": 30,
|
| 28 |
"replay_episode": 0,
|
| 29 |
"push_to_hub": true
|
| 30 |
},
|
| 31 |
-
"mode":
|
| 32 |
-
"device": "
|
| 33 |
}
|
|
|
|
| 7 |
"teleop": null,
|
| 8 |
"processor": {
|
| 9 |
"control_mode": "gamepad",
|
| 10 |
+
"observation": {
|
| 11 |
+
"add_joint_velocity_to_observation": true,
|
| 12 |
+
"display_cameras": false
|
| 13 |
+
},
|
| 14 |
+
"image_preprocessing": null,
|
| 15 |
"gripper": {
|
| 16 |
"use_gripper": true,
|
| 17 |
+
"gripper_penalty": -0.02
|
|
|
|
| 18 |
},
|
| 19 |
"reset": {
|
| 20 |
+
"fixed_reset_joint_positions": [
|
| 21 |
+
0.0,
|
| 22 |
+
0.195,
|
| 23 |
+
0.0,
|
| 24 |
+
-2.43,
|
| 25 |
+
0.0,
|
| 26 |
+
2.62,
|
| 27 |
+
0.785
|
| 28 |
+
],
|
| 29 |
+
"reset_time_s": 1.0,
|
| 30 |
+
"control_time_s": 15.0
|
| 31 |
+
}
|
| 32 |
+
},
|
| 33 |
+
"features": {
|
| 34 |
+
"observation.images.front": {
|
| 35 |
+
"type": "VISUAL",
|
| 36 |
+
"shape": [
|
| 37 |
+
3,
|
| 38 |
+
128,
|
| 39 |
+
128
|
| 40 |
+
]
|
| 41 |
+
},
|
| 42 |
+
"observation.images.wrist": {
|
| 43 |
+
"type": "VISUAL",
|
| 44 |
+
"shape": [
|
| 45 |
+
3,
|
| 46 |
+
128,
|
| 47 |
+
128
|
| 48 |
+
]
|
| 49 |
+
},
|
| 50 |
+
"observation.state": {
|
| 51 |
+
"type": "STATE",
|
| 52 |
+
"shape": [
|
| 53 |
+
18
|
| 54 |
+
]
|
| 55 |
+
},
|
| 56 |
+
"action": {
|
| 57 |
+
"type": "ACTION",
|
| 58 |
+
"shape": [
|
| 59 |
+
3
|
| 60 |
+
]
|
| 61 |
}
|
| 62 |
+
},
|
| 63 |
+
"features_map": {
|
| 64 |
+
"observation.images.front": "observation.images.front",
|
| 65 |
+
"observation.images.wrist": "observation.images.wrist",
|
| 66 |
+
"observation.state": "observation.state",
|
| 67 |
+
"action": "action"
|
| 68 |
}
|
| 69 |
},
|
| 70 |
"dataset": {
|
| 71 |
+
"repo_id": "lilkm/pick_cube_franka_panda_30",
|
| 72 |
"root": null,
|
| 73 |
"task": "PandaPickCubeGamepad-v0",
|
| 74 |
"num_episodes_to_record": 30,
|
| 75 |
"replay_episode": 0,
|
| 76 |
"push_to_hub": true
|
| 77 |
},
|
| 78 |
+
"mode": "record",
|
| 79 |
+
"device": "cuda"
|
| 80 |
}
|
rl/gym_hil/train_config.json
CHANGED
|
@@ -1,75 +1,55 @@
|
|
| 1 |
{
|
| 2 |
"output_dir": null,
|
| 3 |
-
"job_name": "
|
| 4 |
"resume": false,
|
| 5 |
"seed": 1000,
|
| 6 |
"num_workers": 4,
|
| 7 |
"batch_size": 256,
|
| 8 |
"steps": 100000,
|
| 9 |
-
"eval_freq": 20000,
|
| 10 |
"log_freq": 1,
|
| 11 |
"save_checkpoint": true,
|
| 12 |
-
"save_freq":
|
| 13 |
-
"use_policy_training_preset": true,
|
| 14 |
-
"optimizer": null,
|
| 15 |
-
"scheduler": null,
|
| 16 |
"wandb": {
|
| 17 |
"enable": true,
|
| 18 |
"project": "franka_sim",
|
| 19 |
"disable_artifact": true
|
| 20 |
},
|
| 21 |
"dataset": {
|
| 22 |
-
"repo_id": "
|
| 23 |
"use_imagenet_stats": false
|
| 24 |
},
|
| 25 |
-
"
|
| 26 |
-
"type": "
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
"reset": {
|
| 40 |
-
"fixed_reset_joint_positions": [0.0, 0.195, 0.0, -2.43, 0.0, 2.62, 0.785],
|
| 41 |
-
"reset_time_s": 2.0,
|
| 42 |
-
"control_time_s": 15.0,
|
| 43 |
-
"terminate_on_success": true
|
| 44 |
-
}
|
| 45 |
},
|
| 46 |
-
"
|
| 47 |
-
"
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
},
|
| 51 |
-
"observation.images.wrist": {
|
| 52 |
-
"type": "VISUAL",
|
| 53 |
-
"shape": [3, 128, 128]
|
| 54 |
-
},
|
| 55 |
-
"observation.state": {
|
| 56 |
-
"type": "STATE",
|
| 57 |
-
"shape": [18]
|
| 58 |
-
},
|
| 59 |
-
"action": {
|
| 60 |
-
"type": "ACTION",
|
| 61 |
-
"shape": [3]
|
| 62 |
-
}
|
| 63 |
},
|
| 64 |
-
"
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
},
|
|
|
|
|
|
|
| 71 |
"policy": {
|
| 72 |
-
"type": "
|
|
|
|
| 73 |
"n_obs_steps": 1,
|
| 74 |
"normalization_mapping": {
|
| 75 |
"VISUAL": "MEAN_STD",
|
|
@@ -97,8 +77,6 @@
|
|
| 97 |
"shape": [3]
|
| 98 |
}
|
| 99 |
},
|
| 100 |
-
"device": "cuda",
|
| 101 |
-
"use_amp": false,
|
| 102 |
"dataset_stats": {
|
| 103 |
"observation.images.front": {
|
| 104 |
"mean": [0.485, 0.456, 0.406],
|
|
@@ -109,46 +87,66 @@
|
|
| 109 |
"std": [0.229, 0.224, 0.225]
|
| 110 |
},
|
| 111 |
"observation.state": {
|
| 112 |
-
"min": [
|
| 113 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
},
|
| 115 |
"action": {
|
| 116 |
-
"min": [-
|
| 117 |
-
"max": [
|
| 118 |
}
|
| 119 |
},
|
| 120 |
-
"
|
| 121 |
"storage_device": "cpu",
|
| 122 |
-
"vision_encoder_name": "
|
| 123 |
"freeze_vision_encoder": true,
|
| 124 |
"image_encoder_hidden_dim": 32,
|
| 125 |
"shared_encoder": true,
|
| 126 |
"online_steps": 1000000,
|
| 127 |
-
"online_env_seed": 10000,
|
| 128 |
"online_buffer_capacity": 100000,
|
| 129 |
"offline_buffer_capacity": 100000,
|
| 130 |
"online_step_before_learning": 100,
|
| 131 |
-
"policy_update_freq": 1,
|
| 132 |
-
"discount": 0.97,
|
| 133 |
-
"temperature_init": 0.01,
|
| 134 |
-
"num_critics": 2,
|
| 135 |
-
"num_subsample_critics": null,
|
| 136 |
-
"critic_lr": 0.0003,
|
| 137 |
-
"actor_lr": 0.0003,
|
| 138 |
-
"temperature_lr": 0.0003,
|
| 139 |
-
"critic_target_update_weight": 0.005,
|
| 140 |
-
"utd_ratio": 2,
|
| 141 |
"state_encoder_hidden_dim": 256,
|
| 142 |
"latent_dim": 64,
|
| 143 |
-
"
|
| 144 |
-
"
|
| 145 |
-
"grad_clip_norm": 10.0,
|
| 146 |
-
"num_discrete_actions": 3,
|
| 147 |
-
"critic_network_kwargs": {
|
| 148 |
-
"hidden_dims": [256, 256],
|
| 149 |
-
"activate_final": true,
|
| 150 |
-
"final_activation": null
|
| 151 |
-
},
|
| 152 |
"actor_network_kwargs": {
|
| 153 |
"hidden_dims": [256, 256],
|
| 154 |
"activate_final": true
|
|
@@ -159,14 +157,73 @@
|
|
| 159 |
"std_max": 5,
|
| 160 |
"init_final": 0.05
|
| 161 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 162 |
"actor_learner_config": {
|
| 163 |
"learner_host": "127.0.0.1",
|
| 164 |
"learner_port": 50051,
|
| 165 |
-
"policy_parameters_push_frequency":
|
| 166 |
},
|
| 167 |
"concurrency": {
|
| 168 |
"actor": "threads",
|
| 169 |
"learner": "threads"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 170 |
}
|
| 171 |
}
|
| 172 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"output_dir": null,
|
| 3 |
+
"job_name": "default",
|
| 4 |
"resume": false,
|
| 5 |
"seed": 1000,
|
| 6 |
"num_workers": 4,
|
| 7 |
"batch_size": 256,
|
| 8 |
"steps": 100000,
|
|
|
|
| 9 |
"log_freq": 1,
|
| 10 |
"save_checkpoint": true,
|
| 11 |
+
"save_freq": 2000000,
|
|
|
|
|
|
|
|
|
|
| 12 |
"wandb": {
|
| 13 |
"enable": true,
|
| 14 |
"project": "franka_sim",
|
| 15 |
"disable_artifact": true
|
| 16 |
},
|
| 17 |
"dataset": {
|
| 18 |
+
"repo_id": "lilkm/pick_cube_franka_panda_30",
|
| 19 |
"use_imagenet_stats": false
|
| 20 |
},
|
| 21 |
+
"algorithm": {
|
| 22 |
+
"type": "sac",
|
| 23 |
+
"actor_lr": 3e-4,
|
| 24 |
+
"critic_lr": 3e-4,
|
| 25 |
+
"temperature_lr": 3e-4,
|
| 26 |
+
"discount": 0.97,
|
| 27 |
+
"use_backup_entropy": true,
|
| 28 |
+
"critic_target_update_weight": 0.005,
|
| 29 |
+
"num_critics": 2,
|
| 30 |
+
"num_subsample_critics": null,
|
| 31 |
+
"critic_network_kwargs": {
|
| 32 |
+
"hidden_dims": [256, 256],
|
| 33 |
+
"activate_final": true,
|
| 34 |
+
"final_activation": null
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
},
|
| 36 |
+
"discrete_critic_network_kwargs": {
|
| 37 |
+
"hidden_dims": [256, 256],
|
| 38 |
+
"activate_final": true,
|
| 39 |
+
"final_activation": null
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
},
|
| 41 |
+
"temperature_init": 0.01,
|
| 42 |
+
"target_entropy": null,
|
| 43 |
+
"utd_ratio": 2,
|
| 44 |
+
"policy_update_freq": 1,
|
| 45 |
+
"grad_clip_norm": 10.0,
|
| 46 |
+
"use_torch_compile": true
|
| 47 |
},
|
| 48 |
+
"mixer": "online_offline",
|
| 49 |
+
"online_ratio": 0.5,
|
| 50 |
"policy": {
|
| 51 |
+
"type": "gaussian_actor",
|
| 52 |
+
"push_to_hub": false,
|
| 53 |
"n_obs_steps": 1,
|
| 54 |
"normalization_mapping": {
|
| 55 |
"VISUAL": "MEAN_STD",
|
|
|
|
| 77 |
"shape": [3]
|
| 78 |
}
|
| 79 |
},
|
|
|
|
|
|
|
| 80 |
"dataset_stats": {
|
| 81 |
"observation.images.front": {
|
| 82 |
"mean": [0.485, 0.456, 0.406],
|
|
|
|
| 87 |
"std": [0.229, 0.224, 0.225]
|
| 88 |
},
|
| 89 |
"observation.state": {
|
| 90 |
+
"min": [
|
| 91 |
+
-0.18890867,
|
| 92 |
+
-0.13319267,
|
| 93 |
+
-0.15566483,
|
| 94 |
+
-3.04198435,
|
| 95 |
+
-0.05300621,
|
| 96 |
+
2.52864378,
|
| 97 |
+
0.426834,
|
| 98 |
+
-100.0,
|
| 99 |
+
-100.0,
|
| 100 |
+
-100.0,
|
| 101 |
+
-100.0,
|
| 102 |
+
-100.0,
|
| 103 |
+
-100.0,
|
| 104 |
+
-100.0,
|
| 105 |
+
0.0,
|
| 106 |
+
0.2,
|
| 107 |
+
-0.3,
|
| 108 |
+
0.0
|
| 109 |
+
],
|
| 110 |
+
"max": [
|
| 111 |
+
0.1956124,
|
| 112 |
+
0.5096723,
|
| 113 |
+
0.05482981,
|
| 114 |
+
-2.01792539,
|
| 115 |
+
0.04425749,
|
| 116 |
+
2.90565838,
|
| 117 |
+
1.07170398,
|
| 118 |
+
100.0,
|
| 119 |
+
100.0,
|
| 120 |
+
100.0,
|
| 121 |
+
100.0,
|
| 122 |
+
100.0,
|
| 123 |
+
100.0,
|
| 124 |
+
100.0,
|
| 125 |
+
255.0,
|
| 126 |
+
0.6,
|
| 127 |
+
0.3,
|
| 128 |
+
0.5
|
| 129 |
+
]
|
| 130 |
},
|
| 131 |
"action": {
|
| 132 |
+
"min": [-1.0, -1.0, -1.0],
|
| 133 |
+
"max": [1.0, 1.0, 1.0]
|
| 134 |
}
|
| 135 |
},
|
| 136 |
+
"num_discrete_actions": 3,
|
| 137 |
"storage_device": "cpu",
|
| 138 |
+
"vision_encoder_name": "lerobot/resnet10",
|
| 139 |
"freeze_vision_encoder": true,
|
| 140 |
"image_encoder_hidden_dim": 32,
|
| 141 |
"shared_encoder": true,
|
| 142 |
"online_steps": 1000000,
|
|
|
|
| 143 |
"online_buffer_capacity": 100000,
|
| 144 |
"offline_buffer_capacity": 100000,
|
| 145 |
"online_step_before_learning": 100,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 146 |
"state_encoder_hidden_dim": 256,
|
| 147 |
"latent_dim": 64,
|
| 148 |
+
"device": "cuda",
|
| 149 |
+
"use_amp": false,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 150 |
"actor_network_kwargs": {
|
| 151 |
"hidden_dims": [256, 256],
|
| 152 |
"activate_final": true
|
|
|
|
| 157 |
"std_max": 5,
|
| 158 |
"init_final": 0.05
|
| 159 |
},
|
| 160 |
+
"discrete_critic_network_kwargs": {
|
| 161 |
+
"hidden_dims": [256, 256],
|
| 162 |
+
"activate_final": true,
|
| 163 |
+
"final_activation": null
|
| 164 |
+
},
|
| 165 |
"actor_learner_config": {
|
| 166 |
"learner_host": "127.0.0.1",
|
| 167 |
"learner_port": 50051,
|
| 168 |
+
"policy_parameters_push_frequency": 50
|
| 169 |
},
|
| 170 |
"concurrency": {
|
| 171 |
"actor": "threads",
|
| 172 |
"learner": "threads"
|
| 173 |
+
},
|
| 174 |
+
"repo_id": null
|
| 175 |
+
},
|
| 176 |
+
"env": {
|
| 177 |
+
"type": "gym_manipulator",
|
| 178 |
+
"name": "gym_hil",
|
| 179 |
+
"task": "PandaPickCubeGamepad-v0",
|
| 180 |
+
"fps": 10,
|
| 181 |
+
"robot": null,
|
| 182 |
+
"teleop": null,
|
| 183 |
+
"processor": {
|
| 184 |
+
"control_mode": "gamepad",
|
| 185 |
+
"gripper": {
|
| 186 |
+
"use_gripper": true,
|
| 187 |
+
"gripper_penalty": -0.02
|
| 188 |
+
},
|
| 189 |
+
"reset": {
|
| 190 |
+
"fixed_reset_joint_positions": [
|
| 191 |
+
0.0,
|
| 192 |
+
0.195,
|
| 193 |
+
0.0,
|
| 194 |
+
-2.43,
|
| 195 |
+
0.0,
|
| 196 |
+
2.62,
|
| 197 |
+
0.785
|
| 198 |
+
],
|
| 199 |
+
"reset_time_s": 1.0,
|
| 200 |
+
"control_time_s": 15.0
|
| 201 |
+
},
|
| 202 |
+
"image_preprocessing": null
|
| 203 |
+
},
|
| 204 |
+
"features": {
|
| 205 |
+
"observation.images.front": {
|
| 206 |
+
"type": "VISUAL",
|
| 207 |
+
"shape": [3, 128, 128]
|
| 208 |
+
},
|
| 209 |
+
"observation.images.wrist": {
|
| 210 |
+
"type": "VISUAL",
|
| 211 |
+
"shape": [3, 128, 128]
|
| 212 |
+
},
|
| 213 |
+
"observation.state": {
|
| 214 |
+
"type": "STATE",
|
| 215 |
+
"shape": [18]
|
| 216 |
+
},
|
| 217 |
+
"action": {
|
| 218 |
+
"type": "ACTION",
|
| 219 |
+
"shape": [3]
|
| 220 |
+
}
|
| 221 |
+
},
|
| 222 |
+
"features_map": {
|
| 223 |
+
"observation.images.front": "observation.images.front",
|
| 224 |
+
"observation.images.wrist": "observation.images.wrist",
|
| 225 |
+
"observation.state": "observation.state",
|
| 226 |
+
"action": "action"
|
| 227 |
}
|
| 228 |
}
|
| 229 |
}
|
rl/train_config.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"output_dir": null,
|
| 3 |
-
"job_name": "
|
| 4 |
"resume": false,
|
| 5 |
"seed": 1000,
|
| 6 |
"num_workers": 4,
|
|
@@ -8,18 +8,48 @@
|
|
| 8 |
"steps": 100000,
|
| 9 |
"log_freq": 500,
|
| 10 |
"save_checkpoint": true,
|
| 11 |
-
"save_freq":
|
| 12 |
"wandb": {
|
| 13 |
"enable": true,
|
| 14 |
"project": "so100_real",
|
| 15 |
"disable_artifact": true
|
| 16 |
},
|
| 17 |
"dataset": {
|
| 18 |
-
"repo_id": "
|
| 19 |
"use_imagenet_stats": false
|
| 20 |
},
|
| 21 |
-
"
|
| 22 |
"type": "sac",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
"n_obs_steps": 1,
|
| 24 |
"normalization_mapping": {
|
| 25 |
"VISUAL": "MEAN_STD",
|
|
@@ -30,131 +60,57 @@
|
|
| 30 |
"input_features": {
|
| 31 |
"observation.images.top": {
|
| 32 |
"type": "VISUAL",
|
| 33 |
-
"shape": [
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
]
|
| 38 |
},
|
| 39 |
"observation.state": {
|
| 40 |
"type": "STATE",
|
| 41 |
-
"shape": [
|
| 42 |
-
18
|
| 43 |
-
]
|
| 44 |
}
|
| 45 |
},
|
| 46 |
"output_features": {
|
| 47 |
"action": {
|
| 48 |
"type": "ACTION",
|
| 49 |
-
"shape": [
|
| 50 |
-
3
|
| 51 |
-
]
|
| 52 |
}
|
| 53 |
},
|
| 54 |
-
"device": "cuda",
|
| 55 |
-
"use_amp": false,
|
| 56 |
"dataset_stats": {
|
| 57 |
"observation.images.top": {
|
| 58 |
-
"mean": [
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
],
|
| 63 |
-
"std": [
|
| 64 |
-
0.229,
|
| 65 |
-
0.224,
|
| 66 |
-
0.225
|
| 67 |
-
]
|
| 68 |
},
|
| 69 |
"observation.state": {
|
| 70 |
-
"min": [
|
| 71 |
-
|
| 72 |
-
-20.0,
|
| 73 |
-
10.0,
|
| 74 |
-
65.0,
|
| 75 |
-
0.0,
|
| 76 |
-
0.0,
|
| 77 |
-
-40.0,
|
| 78 |
-
-60.0,
|
| 79 |
-
-50.0,
|
| 80 |
-
-80.0,
|
| 81 |
-
-10.0,
|
| 82 |
-
-110.0,
|
| 83 |
-
0.26,
|
| 84 |
-
-0.06,
|
| 85 |
-
0.25
|
| 86 |
-
],
|
| 87 |
-
"max": [
|
| 88 |
-
15.0,
|
| 89 |
-
15.0,
|
| 90 |
-
40.0,
|
| 91 |
-
110.0,
|
| 92 |
-
0.5,
|
| 93 |
-
31.0,
|
| 94 |
-
53.0,
|
| 95 |
-
61.0,
|
| 96 |
-
67.0,
|
| 97 |
-
90.0,
|
| 98 |
-
6.0,
|
| 99 |
-
100.0,
|
| 100 |
-
0.32,
|
| 101 |
-
0.06,
|
| 102 |
-
0.35
|
| 103 |
-
]
|
| 104 |
},
|
| 105 |
"action": {
|
| 106 |
-
"min": [
|
| 107 |
-
|
| 108 |
-
-1.0,
|
| 109 |
-
-1.0
|
| 110 |
-
],
|
| 111 |
-
"max": [
|
| 112 |
-
1.0,
|
| 113 |
-
1.0,
|
| 114 |
-
1.0
|
| 115 |
-
]
|
| 116 |
}
|
| 117 |
},
|
| 118 |
"num_discrete_actions": 3,
|
|
|
|
| 119 |
"storage_device": "cuda",
|
| 120 |
-
"
|
|
|
|
| 121 |
"freeze_vision_encoder": true,
|
| 122 |
"image_encoder_hidden_dim": 32,
|
| 123 |
"shared_encoder": true,
|
| 124 |
"online_steps": 1000000,
|
| 125 |
-
"online_env_seed": 10000,
|
| 126 |
"online_buffer_capacity": 30000,
|
| 127 |
-
"offline_buffer_capacity":
|
| 128 |
-
"online_step_before_learning":
|
| 129 |
-
"policy_update_freq": 1,
|
| 130 |
-
"discount": 0.97,
|
| 131 |
-
"async_prefetch": false,
|
| 132 |
-
"temperature_init": 0.01,
|
| 133 |
-
"num_critics": 2,
|
| 134 |
-
"num_subsample_critics": null,
|
| 135 |
-
"critic_lr": 0.0003,
|
| 136 |
-
"actor_lr": 0.0003,
|
| 137 |
-
"temperature_lr": 0.0003,
|
| 138 |
-
"critic_target_update_weight": 0.005,
|
| 139 |
-
"utd_ratio": 2,
|
| 140 |
"state_encoder_hidden_dim": 256,
|
| 141 |
-
"latent_dim":
|
| 142 |
-
"target_entropy": null,
|
| 143 |
-
"use_backup_entropy": true,
|
| 144 |
-
"grad_clip_norm": 40.0,
|
| 145 |
-
"critic_network_kwargs": {
|
| 146 |
-
"hidden_dims": [
|
| 147 |
-
256,
|
| 148 |
-
256
|
| 149 |
-
],
|
| 150 |
-
"activate_final": true,
|
| 151 |
-
"final_activation": null
|
| 152 |
-
},
|
| 153 |
"actor_network_kwargs": {
|
| 154 |
-
"hidden_dims": [
|
| 155 |
-
256,
|
| 156 |
-
256
|
| 157 |
-
],
|
| 158 |
"activate_final": true
|
| 159 |
},
|
| 160 |
"policy_kwargs": {
|
|
@@ -166,27 +122,40 @@
|
|
| 166 |
"actor_learner_config": {
|
| 167 |
"learner_host": "127.0.0.1",
|
| 168 |
"learner_port": 50051,
|
| 169 |
-
"policy_parameters_push_frequency":
|
| 170 |
},
|
| 171 |
"concurrency": {
|
| 172 |
"actor": "threads",
|
| 173 |
"learner": "threads"
|
| 174 |
},
|
| 175 |
-
"repo_id":
|
| 176 |
},
|
| 177 |
"env": {
|
| 178 |
"type": "gym_manipulator",
|
| 179 |
-
|
|
|
|
|
|
|
|
|
|
| 180 |
"type": "so100_follower",
|
| 181 |
"port": "/dev/ttyACM0",
|
| 182 |
-
"
|
|
|
|
| 183 |
"cameras": {
|
| 184 |
"top": {
|
| 185 |
"type": "opencv",
|
| 186 |
-
"index_or_path":
|
| 187 |
"height": 480,
|
| 188 |
"width": 640,
|
| 189 |
-
"fps": 30
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 190 |
}
|
| 191 |
}
|
| 192 |
},
|
|
@@ -196,90 +165,69 @@
|
|
| 196 |
},
|
| 197 |
"processor": {
|
| 198 |
"control_mode": "gamepad",
|
|
|
|
| 199 |
"observation": {
|
| 200 |
"display_cameras": false,
|
| 201 |
"add_joint_velocity_to_observation": true,
|
| 202 |
"add_current_to_observation": true,
|
| 203 |
-
"add_ee_pose_to_observation":
|
| 204 |
},
|
| 205 |
"image_preprocessing": {
|
| 206 |
-
"crop_params_dict":
|
| 207 |
-
"observation.images.top": [49, 189, 262, 305]
|
| 208 |
-
},
|
| 209 |
"resize_size": [128, 128]
|
| 210 |
},
|
| 211 |
"gripper": {
|
| 212 |
"use_gripper": true,
|
| 213 |
-
"gripper_penalty": -0.02
|
| 214 |
-
"gripper_penalty_in_reward": false
|
| 215 |
},
|
| 216 |
"reset": {
|
| 217 |
-
"fixed_reset_joint_positions": [
|
| 218 |
-
0.0,
|
| 219 |
-
0.0,
|
| 220 |
-
0.0,
|
| 221 |
-
90.0,
|
| 222 |
-
0.0,
|
| 223 |
-
30.0
|
| 224 |
-
],
|
| 225 |
"reset_time_s": 2.5,
|
| 226 |
-
"control_time_s":
|
| 227 |
"terminate_on_success": true
|
| 228 |
},
|
| 229 |
"inverse_kinematics": {
|
| 230 |
-
"urdf_path": "/
|
| 231 |
"target_frame_name": "gripper_frame_link",
|
| 232 |
"end_effector_bounds": {
|
| 233 |
-
"min": [
|
| 234 |
-
|
| 235 |
-
-0.065,
|
| 236 |
-
0.005
|
| 237 |
-
],
|
| 238 |
-
"max": [
|
| 239 |
-
0.25,
|
| 240 |
-
0.065,
|
| 241 |
-
0.1
|
| 242 |
-
]
|
| 243 |
},
|
| 244 |
"end_effector_step_sizes": {
|
| 245 |
-
"x": 0.
|
| 246 |
-
"y": 0.
|
| 247 |
-
"z": 0.
|
| 248 |
}
|
| 249 |
},
|
| 250 |
"reward_classifier": {
|
| 251 |
"pretrained_path": null,
|
| 252 |
"success_threshold": 0.5,
|
| 253 |
"success_reward": 1.0
|
| 254 |
-
}
|
| 255 |
-
"max_gripper_pos": 30
|
| 256 |
},
|
| 257 |
-
"name": "real_robot",
|
| 258 |
-
"fps": 10,
|
| 259 |
-
|
| 260 |
-
"task": "",
|
| 261 |
"features": {
|
| 262 |
"observation.images.top": {
|
| 263 |
"type": "VISUAL",
|
| 264 |
-
"shape": [
|
| 265 |
-
|
| 266 |
-
|
| 267 |
-
|
| 268 |
-
]
|
| 269 |
},
|
| 270 |
"observation.state": {
|
| 271 |
"type": "STATE",
|
| 272 |
-
"shape": [
|
| 273 |
-
18
|
| 274 |
-
]
|
| 275 |
},
|
| 276 |
"action": {
|
| 277 |
"type": "ACTION",
|
| 278 |
-
"shape": [
|
| 279 |
-
3
|
| 280 |
-
]
|
| 281 |
}
|
| 282 |
},
|
| 283 |
-
"features_map": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 284 |
}
|
| 285 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"output_dir": null,
|
| 3 |
+
"job_name": "so100_hilserl",
|
| 4 |
"resume": false,
|
| 5 |
"seed": 1000,
|
| 6 |
"num_workers": 4,
|
|
|
|
| 8 |
"steps": 100000,
|
| 9 |
"log_freq": 500,
|
| 10 |
"save_checkpoint": true,
|
| 11 |
+
"save_freq": 5000,
|
| 12 |
"wandb": {
|
| 13 |
"enable": true,
|
| 14 |
"project": "so100_real",
|
| 15 |
"disable_artifact": true
|
| 16 |
},
|
| 17 |
"dataset": {
|
| 18 |
+
"repo_id": "lilkm/so100_gamepad_insert_high_quality",
|
| 19 |
"use_imagenet_stats": false
|
| 20 |
},
|
| 21 |
+
"algorithm": {
|
| 22 |
"type": "sac",
|
| 23 |
+
"actor_lr": 3e-4,
|
| 24 |
+
"critic_lr": 3e-4,
|
| 25 |
+
"temperature_lr": 3e-4,
|
| 26 |
+
"discount": 0.97,
|
| 27 |
+
"use_backup_entropy": true,
|
| 28 |
+
"critic_target_update_weight": 0.005,
|
| 29 |
+
"num_critics": 2,
|
| 30 |
+
"num_subsample_critics": null,
|
| 31 |
+
"critic_network_kwargs": {
|
| 32 |
+
"hidden_dims": [256, 256],
|
| 33 |
+
"activate_final": true,
|
| 34 |
+
"final_activation": null
|
| 35 |
+
},
|
| 36 |
+
"discrete_critic_network_kwargs": {
|
| 37 |
+
"hidden_dims": [256, 256],
|
| 38 |
+
"activate_final": true,
|
| 39 |
+
"final_activation": null
|
| 40 |
+
},
|
| 41 |
+
"temperature_init": 0.01,
|
| 42 |
+
"target_entropy": null,
|
| 43 |
+
"utd_ratio": 2,
|
| 44 |
+
"policy_update_freq": 1,
|
| 45 |
+
"grad_clip_norm": 10.0,
|
| 46 |
+
"use_torch_compile": false
|
| 47 |
+
},
|
| 48 |
+
"mixer": "online_offline",
|
| 49 |
+
"online_ratio": 0.5,
|
| 50 |
+
"policy": {
|
| 51 |
+
"type": "gaussian_actor",
|
| 52 |
+
"push_to_hub": false,
|
| 53 |
"n_obs_steps": 1,
|
| 54 |
"normalization_mapping": {
|
| 55 |
"VISUAL": "MEAN_STD",
|
|
|
|
| 60 |
"input_features": {
|
| 61 |
"observation.images.top": {
|
| 62 |
"type": "VISUAL",
|
| 63 |
+
"shape": [3, 128, 128]
|
| 64 |
+
},
|
| 65 |
+
"observation.images.wrist": {
|
| 66 |
+
"type": "VISUAL",
|
| 67 |
+
"shape": [3, 128, 128]
|
| 68 |
},
|
| 69 |
"observation.state": {
|
| 70 |
"type": "STATE",
|
| 71 |
+
"shape": [18]
|
|
|
|
|
|
|
| 72 |
}
|
| 73 |
},
|
| 74 |
"output_features": {
|
| 75 |
"action": {
|
| 76 |
"type": "ACTION",
|
| 77 |
+
"shape": [3]
|
|
|
|
|
|
|
| 78 |
}
|
| 79 |
},
|
|
|
|
|
|
|
| 80 |
"dataset_stats": {
|
| 81 |
"observation.images.top": {
|
| 82 |
+
"mean": [0.485, 0.456, 0.406],
|
| 83 |
+
"std": [0.229, 0.224, 0.225]
|
| 84 |
+
},
|
| 85 |
+
"observation.images.wrist": {
|
| 86 |
+
"mean": [0.485, 0.456, 0.406],
|
| 87 |
+
"std": [0.229, 0.224, 0.225]
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
},
|
| 89 |
"observation.state": {
|
| 90 |
+
"min": [-180, -180, -180, -180, -180, -180, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100, -100],
|
| 91 |
+
"max": [180, 180, 180, 180, 180, 180, 100, 100, 100, 100, 100, 100, 100, 100, 100, 100, 100, 100]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 92 |
},
|
| 93 |
"action": {
|
| 94 |
+
"min": [-1.0, -1.0, -1.0],
|
| 95 |
+
"max": [1.0, 1.0, 1.0]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
}
|
| 97 |
},
|
| 98 |
"num_discrete_actions": 3,
|
| 99 |
+
"device": "cuda",
|
| 100 |
"storage_device": "cuda",
|
| 101 |
+
"use_amp": false,
|
| 102 |
+
"vision_encoder_name": "lerobot/resnet10",
|
| 103 |
"freeze_vision_encoder": true,
|
| 104 |
"image_encoder_hidden_dim": 32,
|
| 105 |
"shared_encoder": true,
|
| 106 |
"online_steps": 1000000,
|
|
|
|
| 107 |
"online_buffer_capacity": 30000,
|
| 108 |
+
"offline_buffer_capacity": 20000,
|
| 109 |
+
"online_step_before_learning": 100,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 110 |
"state_encoder_hidden_dim": 256,
|
| 111 |
+
"latent_dim": 64,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 112 |
"actor_network_kwargs": {
|
| 113 |
+
"hidden_dims": [256, 256],
|
|
|
|
|
|
|
|
|
|
| 114 |
"activate_final": true
|
| 115 |
},
|
| 116 |
"policy_kwargs": {
|
|
|
|
| 122 |
"actor_learner_config": {
|
| 123 |
"learner_host": "127.0.0.1",
|
| 124 |
"learner_port": 50051,
|
| 125 |
+
"policy_parameters_push_frequency": 50
|
| 126 |
},
|
| 127 |
"concurrency": {
|
| 128 |
"actor": "threads",
|
| 129 |
"learner": "threads"
|
| 130 |
},
|
| 131 |
+
"repo_id": null
|
| 132 |
},
|
| 133 |
"env": {
|
| 134 |
"type": "gym_manipulator",
|
| 135 |
+
"name": "real_robot",
|
| 136 |
+
"fps": 10,
|
| 137 |
+
"task": "usb insert",
|
| 138 |
+
"robot": {
|
| 139 |
"type": "so100_follower",
|
| 140 |
"port": "/dev/ttyACM0",
|
| 141 |
+
"use_degrees": true,
|
| 142 |
+
"id": "blue_follower",
|
| 143 |
"cameras": {
|
| 144 |
"top": {
|
| 145 |
"type": "opencv",
|
| 146 |
+
"index_or_path": "/dev/video2",
|
| 147 |
"height": 480,
|
| 148 |
"width": 640,
|
| 149 |
+
"fps": 30,
|
| 150 |
+
"fourcc": "MJPG"
|
| 151 |
+
},
|
| 152 |
+
"wrist": {
|
| 153 |
+
"type": "opencv",
|
| 154 |
+
"index_or_path": "/dev/video4",
|
| 155 |
+
"height": 480,
|
| 156 |
+
"width": 640,
|
| 157 |
+
"fps": 30,
|
| 158 |
+
"fourcc": "MJPG"
|
| 159 |
}
|
| 160 |
}
|
| 161 |
},
|
|
|
|
| 165 |
},
|
| 166 |
"processor": {
|
| 167 |
"control_mode": "gamepad",
|
| 168 |
+
"max_gripper_pos": 30,
|
| 169 |
"observation": {
|
| 170 |
"display_cameras": false,
|
| 171 |
"add_joint_velocity_to_observation": true,
|
| 172 |
"add_current_to_observation": true,
|
| 173 |
+
"add_ee_pose_to_observation": false
|
| 174 |
},
|
| 175 |
"image_preprocessing": {
|
| 176 |
+
"crop_params_dict": null,
|
|
|
|
|
|
|
| 177 |
"resize_size": [128, 128]
|
| 178 |
},
|
| 179 |
"gripper": {
|
| 180 |
"use_gripper": true,
|
| 181 |
+
"gripper_penalty": -0.02
|
|
|
|
| 182 |
},
|
| 183 |
"reset": {
|
| 184 |
+
"fixed_reset_joint_positions": [0.0, -20.0, 25.0, 90.0, 90.0, 30.0],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 185 |
"reset_time_s": 2.5,
|
| 186 |
+
"control_time_s": 30.0,
|
| 187 |
"terminate_on_success": true
|
| 188 |
},
|
| 189 |
"inverse_kinematics": {
|
| 190 |
+
"urdf_path": "./SO101/so101_new_calib.urdf",
|
| 191 |
"target_frame_name": "gripper_frame_link",
|
| 192 |
"end_effector_bounds": {
|
| 193 |
+
"min": [0.115, -0.145, -0.0075],
|
| 194 |
+
"max": [0.26, 0.145, 0.1]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 195 |
},
|
| 196 |
"end_effector_step_sizes": {
|
| 197 |
+
"x": 0.009,
|
| 198 |
+
"y": 0.009,
|
| 199 |
+
"z": 0.01
|
| 200 |
}
|
| 201 |
},
|
| 202 |
"reward_classifier": {
|
| 203 |
"pretrained_path": null,
|
| 204 |
"success_threshold": 0.5,
|
| 205 |
"success_reward": 1.0
|
| 206 |
+
}
|
|
|
|
| 207 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
| 208 |
"features": {
|
| 209 |
"observation.images.top": {
|
| 210 |
"type": "VISUAL",
|
| 211 |
+
"shape": [3, 128, 128]
|
| 212 |
+
},
|
| 213 |
+
"observation.images.wrist": {
|
| 214 |
+
"type": "VISUAL",
|
| 215 |
+
"shape": [3, 128, 128]
|
| 216 |
},
|
| 217 |
"observation.state": {
|
| 218 |
"type": "STATE",
|
| 219 |
+
"shape": [18]
|
|
|
|
|
|
|
| 220 |
},
|
| 221 |
"action": {
|
| 222 |
"type": "ACTION",
|
| 223 |
+
"shape": [3]
|
|
|
|
|
|
|
| 224 |
}
|
| 225 |
},
|
| 226 |
+
"features_map": {
|
| 227 |
+
"observation.images.top": "observation.images.top",
|
| 228 |
+
"observation.images.wrist": "observation.images.wrist",
|
| 229 |
+
"observation.state": "observation.state",
|
| 230 |
+
"action": "action"
|
| 231 |
+
}
|
| 232 |
}
|
| 233 |
}
|