Add pushT env

configs/agents/rl/push_t/gym_config.json

@@ -0,0 +1,193 @@
+{ 
+    "id": "PushTRL",
+    "max_episodes": 5,
+    "env": {
+        "num_envs": 128,
+        "sim_steps_per_control": 4,
+        "events": {
+            "randomize_t": {
+                "func": "randomize_rigid_object_pose",
+                "mode": "reset",
+                "params": {
+                    "entity_cfg": {
+                        "uid": "t"
+                    },
+                    "position_range": [
+                        [-0.2, -0.2, 0.0],
+                        [0.2, 0.2, 0.0]
+                    ],
+                    "relative_position": true
+                }
+            },
+            "randomize_goal": {
+                "func": "randomize_target_pose",
+                "mode": "reset",
+                "params": {
+                    "position_range": [
+                        [-0.3, -0.3, 0.05],
+                        [0.3, 0.3, 0.05]
+                    ],
+                    "relative_position": false,
+                    "store_key": "goal_pose"
+                }
+            }
+        },
+        "observations": {
+            "robot_qpos": {
+                "func": "normalize_robot_joint_data",
+                "mode": "modify",
+                "name": "robot/qpos",
+                "params": {
+                    "joint_ids": [0, 1, 2, 3, 4, 5]
+                }
+            },
+            "robot_ee_pos": {
+                "func": "get_robot_eef_pose",
+                "mode": "add",
+                "name": "robot/ee_pos",
+                "params": {
+                    "part_name": "arm"
+                }
+            },
+            "t_pos": {
+                "func": "get_rigid_object_pose",
+                "mode": "add",
+                "name": "object/t_pos",
+                "params": {
+                    "entity_cfg": {
+                        "uid": "t"
+                    }
+                }
+            },
+            "goal_pos": {
+                "func": "target_position",
+                "mode": "add",
+                "name": "object/goal_pos",
+                "params": {
+                    "target_pose_key": "goal_pose"
+                }
+            }
+        },
+        "rewards": {
+            "reaching_reward": {
+                "func": "reaching_behind_object",
+                "mode": "add",
+                "weight": 0.1,
+                "params": {
+                    "object_cfg": {
+                        "uid": "t"
+                    },
+                    "target_pose_key": "goal_pose",
+                    "behind_offset": 0.015,
+                    "height_offset": 0.015,
+                    "distance_scale": 5.0,
+                    "part_name": "arm"
+                }
+            },
+            "place_reward": {
+                "func": "incremental_distance_to_target",
+                "mode": "add",
+                "weight": 1.0,
+                "params": {
+                    "source_entity_cfg": {
+                        "uid": "t"
+                    },
+                    "target_pose_key": "goal_pose",
+                    "tanh_scale": 10.0,
+                    "positive_weight": 2.0,
+                    "negative_weight": 0.5,
+                    "use_xy_only": true
+                }
+            },
+            "action_penalty": {
+                "func": "action_smoothness_penalty",
+                "mode": "add",
+                "weight": 0.01,
+                "params": {}
+            },
+            "success_bonus": {
+                "func": "success_reward",
+                "mode": "add",
+                "weight": 10.0,
+                "params": {}
+            }
+        },
+        "extensions": {
+            "obs_mode": "state",
+            "episode_length": 100,
+            "joint_limits": 0.5,
+            "action_scale": 0.1,
+            "success_threshold": 0.1
+        }
+    },
+    "robot": {
+        "uid": "Manipulator",
+        "urdf_cfg": {
+            "components": [
+                {
+                    "component_type": "arm",
+                    "urdf_path": "UniversalRobots/UR10/UR10.urdf"
+                },
+                {
+                    "component_type": "hand",
+                    "urdf_path": "DH_PGI_140_80/DH_PGI_140_80.urdf"
+                }
+            ]
+        },
+        "init_pos": [0.0, 0.0, 0.0],
+        "init_rot": [0.0, 0.0, 0.0],
+        "init_qpos": [0.0, -1.57, 1.57, -1.57, -1.57, 0.0, 0.04, 0.04],
+        "drive_pros": {
+            "drive_type": "force",
+            "stiffness": 100000.0,
+            "damping": 1000.0,
+            "max_velocity": 2.0,
+            "max_effort": 500.0
+        },
+        "solver_cfg": {
+            "arm": {
+                "class_type": "PytorchSolver",
+                "end_link_name": "ee_link",
+                "root_link_name": "base_link",
+                "tcp": [
+                    [1.0, 0.0, 0.0, 0.0],
+                    [0.0, 1.0, 0.0, 0.0],
+                    [0.0, 0.0, 1.0, 0.16],
+                    [0.0, 0.0, 0.0, 1.0]
+                ]
+            }
+        },
+        "control_parts": {
+            "arm": ["JOINT[1-6]"]
+        }
+    },
+    "sensor": [],
+    "light": {},
+    "background": [],
+    "rigid_object": [
+        {
+            "uid": "t",
+            "shape": {
+                "shape_type": "Cube",
+                "size": [0.1, 0.1, 0.1]
+            },
+            "body_type": "dynamic",
+            "init_pos": [-0.6, -0.4, 0.05],
+            "attrs": {
+                "mass": 10.0,
+                "static_friction": 3.0,
+                "dynamic_friction": 2.0,
+                "linear_damping": 2.0,
+                "angular_damping": 2.0,
+                "contact_offset": 0.003,
+                "rest_offset": 0.001,
+                "restitution": 0.1,
+                "max_depenetration_velocity": 10.0,
+                "max_linear_velocity": 1.0,
+                "max_angular_velocity": 1.0
+            }
+        }
+    ],
+    "rigid_object_group": [],
+    "articulation": []
+}

configs/agents/rl/push_t/train_config.json

@@ -0,0 +1,67 @@
+{ 
+    "trainer": {
+        "exp_name": "push_t_ppo",
+        "gym_config": "configs/agents/rl/push_t/gym_config.json",
+        "seed": 42,
+        "device": "cuda:0",
+        "headless": false,
+        "enable_rt": false,
+        "gpu_id": 0,
+        "num_envs": 8,
+        "iterations": 1000,
+        "rollout_steps": 1024,
+        "eval_freq": 200,
+        "save_freq": 200,
+        "use_wandb": true,
+        "wandb_project_name": "embodychain-push_t",
+        "events": {
+            "eval": {
+                "record_camera": {
+                    "func": "record_camera_data_async",
+                    "mode": "interval",
+                    "interval_step": 1,
+                    "params": {
+                        "name": "main_cam",
+                        "resolution": [640, 480],
+                        "eye": [-1.4, 1.4, 2.0],
+                        "target": [0, 0, 0],
+                        "up": [0, 0, 1],
+                        "intrinsics": [600, 600, 320, 240],
+                        "save_path": "./outputs/videos/eval"
+                    }
+                }
+            }
+        }
+    },
+    "policy": {
+        "name": "actor_critic",
+        "actor": {
+            "type": "mlp",
+            "network_cfg": {
+                "hidden_sizes": [256, 256],
+                "activation": "relu"
+            }
+        },
+        "critic": {
+            "type": "mlp",
+            "network_cfg": {
+                "hidden_sizes": [256, 256],
+                "activation": "relu"
+            }
+        }
+    },
+    "algorithm": {
+        "name": "ppo",
+        "cfg": {
+            "learning_rate": 0.0001,
+            "n_epochs": 10,
+            "batch_size": 8192,
+            "gamma": 0.99,
+            "gae_lambda": 0.95,
+            "clip_coef": 0.2,
+            "ent_coef": 0.01,
+            "vf_coef": 0.5,
+            "max_grad_norm": 0.5
+        }
+    }
+}