Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,58 @@
+---
+license: apache-2.0
+task_categories:
+- reinforcement-learning
+- robotics
+tags:
+- Minari
+- TeleArms
+- bimanual
+- simulation
+- offline-rl
+---
+# TeleArms offline RL datasets (state)
+
+30 [Minari](https://minari.farama.org) datasets of human teleoperated demonstrations of 10 bimanual
+YAM cube tasks in MuJoCo: each task in three subsets of 1,500 demos, picked at random
+(`random-1500`), by player ELO (`elo-1500`) and by demo score (`score-1500`). The ids are
+`telearms/<Task>/<subset>-v0`:
+
+```python
+import minari
+minari.download_dataset("hf://mliu48/telearms/CubePatchPlacement/random-1500-v0")
+dataset = minari.load_dataset("telearms/CubePatchPlacement/random-1500-v0")
+env = dataset.recover_environment()  # needs training/telearms_env.py on the path
+```
+
+The environment is `TeleArmsEnv(task, obs_mode="privileged")` in
+[training/telearms_env.py](https://github.com/helen9975/RoboEval-telearms/tree/gymnasium-env). On a machine without EGL, set `MUJOCO_GL=disable`
+before importing it.
+
+**Observations** are the simulator's state, not images: the robot's 14 joint and gripper values,
+then each cube's position, orientation and velocity (13 numbers per cube), then each goal patch's
+position and orientation (7 per patch).
+
+| Task | Observation size |
+|---|---|
+| CubePatchPlacement | 54 |
+| FrontEdgeCubeRow | 53 |
+| FrontToBackCubeLine | 53 |
+| LCubeArrangement | 66 |
+| LetterFaceAlignment | 27 |
+| RedCubeInFrontOfBlueCube | 40 |
+| RedLeftOfBluePlacement | 40 |
+| SemanticCubeOnRedPatchAUp | 34 |
+| SequentialPushBlockToTarget | 54 |
+| ThreeCubeWallRow | 53 |
+
+**Rewards.** Every demo was replayed in its task's environment and scored by its rules. `rewards`
+is the sparse reward, 1.0 on the step that reaches success, else 0. `infos["reward_dense"]` is the
+environment's `reward_type="dense"` reward: 1/N on the first step that reaches each of the task's
+N stages (grasp, lift, place, ...) plus 1.0 on success, at most 2 per episode. Each demo ends at
+its first success (`terminations`), as the environment ends an episode. Of the 45,000 demos,
+1.5% were accepted when played but never satisfy the current task check; they end with
+`truncations` and no reward.
+
+The same demos with camera images are the `offline-rl` branch of
+`mliu48/telearms-yam-cube-v5-exp-<subset>`:
+`LeRobotDataset("mliu48/telearms-yam-cube-v5-exp-random-1500", revision="offline-rl")`.