mirror of
https://github.com/huggingface/lerobot.git
synced 2026-05-16 00:59:46 +00:00
Pepijn
49186359b0
Mirror the libero/metaworld/libero_plus/robomme pattern: start from the nightly GPU image (apt deps, python, uv, venv, lerobot[all] already there) and layer on only what RoboTwin 2.0 uniquely needs — cuda-nvcc + cuda-cudart-dev (CuRobo builds from source), Vulkan libs + NVIDIA ICD (SAPIEN renderer), sapien/mplib/open3d/pytorch3d/curobo installs, the mplib + sapien upstream patches, and the TianxingChen asset download. Drops ~90 lines of duplicated base setup (CUDA FROM, apt python, uv install, user creation, venv init, base lerobot install). 199 → 110. Also repoint the docs + env docstring dataset link from hxma/RoboTwin-LeRobot-v3.0 to the canonical lerobot/robotwin_unified. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
207 lines
8.3 KiB
Plaintext
207 lines
8.3 KiB
Plaintext
# RoboTwin 2.0
|
|
|
|
RoboTwin 2.0 is a **large-scale dual-arm manipulation benchmark** built on the SAPIEN physics engine. It provides a standardized evaluation protocol for bimanual robotic policies across 60 tasks with strong domain randomization (clutter, lighting, background, tabletop height, and language instructions).
|
|
|
|
- Paper: [RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation](https://robotwin-platform.github.io)
|
|
- GitHub: [RoboTwin-Platform/RoboTwin](https://github.com/RoboTwin-Platform/RoboTwin)
|
|
- Leaderboard: [robotwin-platform.github.io/leaderboard](https://robotwin-platform.github.io/leaderboard)
|
|
- Dataset: [lerobot/robotwin_unified](https://huggingface.co/datasets/lerobot/robotwin_unified)
|
|
|
|
## Overview
|
|
|
|
| Property | Value |
|
|
| ------------- | ---------------------------------------------------------- |
|
|
| Tasks | 60 dual-arm manipulation tasks |
|
|
| Robot | Aloha-AgileX bimanual (14 DOF, 7 per arm) |
|
|
| Action space | 14-dim joint-space, continuous in `[-1, 1]` |
|
|
| Cameras | `head_camera`, `front_camera`, `left_wrist`, `right_wrist` |
|
|
| Simulator | SAPIEN (not MuJoCo) |
|
|
| Eval protocol | 100 episodes/task, 50 demo_clean demonstrations |
|
|
| Eval settings | **Easy** (`demo_clean`) and **Hard** (`demo_randomized`) |
|
|
|
|
## Available tasks
|
|
|
|
RoboTwin 2.0 ships with 60 dual-arm manipulation tasks. The full list appears on the [leaderboard](https://robotwin-platform.github.io/leaderboard). Example tasks:
|
|
|
|
| Task | CLI name | Category |
|
|
| ---------------------- | ------------------------ | ---------------- |
|
|
| Beat block with hammer | `beat_block_hammer` | Tool use |
|
|
| Open / close laptop | `open_laptop` | Articulated obj |
|
|
| Stack blocks (2 / 3) | `stack_blocks_two/three` | Stacking |
|
|
| Pour water | `pour_water` | Deformable/fluid |
|
|
| Fold cloth | `fold_cloth` | Deformable |
|
|
| Handover block | `handover_block` | Bimanual coord. |
|
|
| Place shoes | `place_shoes_left/right` | Precision place |
|
|
| Scan object | `scan_object` | Mobile manip. |
|
|
|
|
Pass a comma-separated list to `--env.task` to run multiple tasks in a single eval sweep.
|
|
|
|
## Dataset
|
|
|
|
The RoboTwin 2.0 dataset is available in **LeRobot v3.0 format** on the Hugging Face Hub:
|
|
|
|
```
|
|
lerobot/robotwin_unified
|
|
```
|
|
|
|
It contains over 100,000 pre-collected trajectories across all 60 tasks (79.6 GB, Apache 2.0 license). No format conversion is needed — it is already in the correct LeRobot v3.0 schema with video observations and action labels.
|
|
|
|
You can load it directly with the HF Datasets library:
|
|
|
|
```python
|
|
from datasets import load_dataset
|
|
|
|
ds = load_dataset("lerobot/robotwin_unified", split="train")
|
|
```
|
|
|
|
## Installation
|
|
|
|
RoboTwin 2.0 requires **Linux** with an NVIDIA GPU (CUDA 12.1 recommended). Installation takes approximately 20 minutes.
|
|
|
|
### 1. Create a conda environment
|
|
|
|
```bash
|
|
conda create -n robotwin python=3.10 -y
|
|
conda activate robotwin
|
|
```
|
|
|
|
### 2. Install LeRobot
|
|
|
|
```bash
|
|
git clone https://github.com/huggingface/lerobot.git
|
|
cd lerobot
|
|
pip install -e "."
|
|
```
|
|
|
|
### 3. Install RoboTwin 2.0
|
|
|
|
```bash
|
|
git clone https://github.com/RoboTwin-Platform/RoboTwin.git
|
|
cd RoboTwin
|
|
bash script/_install.sh
|
|
bash script/_download_assets.sh
|
|
```
|
|
|
|
The install script handles all Python dependencies including SAPIEN, CuRobo, mplib, and pytorch3d.
|
|
|
|
<Tip warning={true}>
|
|
If the automated install fails, install manually:
|
|
|
|
```bash
|
|
pip install -r requirements.txt
|
|
pip install "git+https://github.com/facebookresearch/pytorch3d.git@stable"
|
|
cd envs && git clone https://github.com/NVlabs/curobo.git && cd curobo
|
|
pip install -e . --no-build-isolation
|
|
```
|
|
|
|
Then apply the required mplib fix: in `mplib/planner.py` line 807, remove `or collide` from the conditional.
|
|
|
|
</Tip>
|
|
|
|
### 4. Add RoboTwin to PYTHONPATH
|
|
|
|
The RoboTwin task modules must be importable by LeRobot. From within the `RoboTwin/` directory:
|
|
|
|
```bash
|
|
export PYTHONPATH="${PYTHONPATH}:$(pwd)"
|
|
```
|
|
|
|
Add this to your shell profile to make it permanent.
|
|
|
|
## Evaluation
|
|
|
|
### Standard evaluation (recommended)
|
|
|
|
Evaluate a policy on a single task with the official protocol (100 episodes):
|
|
|
|
```bash
|
|
lerobot-eval \
|
|
--policy.path="your-hf-policy-id" \
|
|
--env.type=robotwin \
|
|
--env.task=beat_block_hammer \
|
|
--eval.batch_size=1 \
|
|
--eval.n_episodes=100
|
|
```
|
|
|
|
### Single-task quick check
|
|
|
|
```bash
|
|
lerobot-eval \
|
|
--policy.path="your-hf-policy-id" \
|
|
--env.type=robotwin \
|
|
--env.task=beat_block_hammer \
|
|
--eval.batch_size=1 \
|
|
--eval.n_episodes=5
|
|
```
|
|
|
|
### Multi-task sweep
|
|
|
|
Evaluate on several tasks in one run:
|
|
|
|
```bash
|
|
lerobot-eval \
|
|
--policy.path="your-hf-policy-id" \
|
|
--env.type=robotwin \
|
|
--env.task=beat_block_hammer,click_bell,handover_block,stack_blocks_two \
|
|
--eval.batch_size=1 \
|
|
--eval.n_episodes=100
|
|
```
|
|
|
|
### Full benchmark (all 60 tasks)
|
|
|
|
```bash
|
|
lerobot-eval \
|
|
--policy.path="your-hf-policy-id" \
|
|
--env.type=robotwin \
|
|
--env.task=adjust_bottle,beat_block_hammer,blocks_ranking_rgb,blocks_ranking_size,click_alarmclock,click_bell,close_laptop,close_microwave,dump_bin,grab_roller,handover_block,handover_cup,handover_diverse_bottles,handover_mic,hanging_mug,insert_pin,lift_pot,make_tea,open_laptop,open_microwave,pick_diverse_bottles,pick_dual_bottles,place_basket,place_block,place_cable,place_can,place_chopsticks,place_cloth,place_container,place_cup,place_diverse_bottles,place_dual_bottles,place_fork,place_knife,place_object_basket,place_ring,place_ruler,place_shoes_left,place_shoes_right,place_spoon,place_toy,pour_water,press_stapler,put_bottles_dustbin,put_object_cabinet,put_shoes_box,rotate_qrcode,scan_object,shake_bottle,shake_bottle_horizontally,stack_blocks_three,stack_blocks_two,stack_bowls_three,stack_bowls_two,stamp_seal,turn_switch,wipe_board,arrange_tools,build_tower,fold_cloth \
|
|
--eval.batch_size=1 \
|
|
--eval.n_episodes=100
|
|
```
|
|
|
|
## Camera configuration
|
|
|
|
By default, all four cameras are included:
|
|
|
|
| Camera key | Description |
|
|
| -------------- | ------------------------------ |
|
|
| `head_camera` | Overhead / third-person view |
|
|
| `front_camera` | Front-facing static camera |
|
|
| `left_wrist` | Left arm wrist-mounted camera |
|
|
| `right_wrist` | Right arm wrist-mounted camera |
|
|
|
|
To use a subset of cameras, override `--env.camera_names`:
|
|
|
|
```bash
|
|
lerobot-eval \
|
|
--policy.path="your-hf-policy-id" \
|
|
--env.type=robotwin \
|
|
--env.task=beat_block_hammer \
|
|
--env.camera_names="head_camera,left_wrist,right_wrist" \
|
|
--eval.batch_size=1 \
|
|
--eval.n_episodes=10
|
|
```
|
|
|
|
## Environment config reference
|
|
|
|
Key parameters for `RoboTwinEnvConfig`:
|
|
|
|
| Parameter | Default | Description |
|
|
| -------------------- | --------------------------------------------------- | ---------------------------------- |
|
|
| `task` | `"beat_block_hammer"` | Comma-separated task name(s) |
|
|
| `fps` | `25` | Simulation FPS |
|
|
| `episode_length` | `300` | Max steps per episode |
|
|
| `obs_type` | `"pixels_agent_pos"` | `"pixels"` or `"pixels_agent_pos"` |
|
|
| `camera_names` | `"head_camera,front_camera,left_wrist,right_wrist"` | Comma-separated active cameras |
|
|
| `observation_height` | `480` | Camera pixel height |
|
|
| `observation_width` | `640` | Camera pixel width |
|
|
|
|
## Leaderboard submission
|
|
|
|
Results can be submitted to the [RoboTwin 2.0 leaderboard](https://robotwin-platform.github.io/leaderboard). The official protocol requires:
|
|
|
|
- Training on 50 `demo_clean` demonstrations per task
|
|
- Evaluating 100 episodes per task
|
|
- Reporting success rate separately for **Easy** (`demo_clean`) and **Hard** (`demo_randomized`) settings
|
|
|
|
For submission instructions, refer to the [RoboTwin 2.0 documentation](https://robotwin-platform.github.io/doc/).
|