> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/NVIDIA/Isaac-GR00T/llms.txt
> Use this file to discover all available pages before exploring further.

# Embodiment tags

> Robot embodiment identifiers for cross-embodiment training

Embodiment tags are used to identify the robot embodiment in the data. They enable cross-embodiment training by providing consistent identifiers across datasets.

## Naming convention

Embodiment tags follow the pattern `<dataset>_<robot_name>`. If using multiple datasets with the same robot (e.g., sim GR1 and real GR1), you can drop the dataset name and use only the robot name.

## EmbodimentTag

```python theme={null}
from gr00t.data.embodiment_tags import EmbodimentTag
```

Enum class for robot embodiment identifiers.

### Pretrain embodiment tags

These embodiments are used during the pretraining phase:

#### ROBOCASA\_PANDA\_OMRON

The RoboCasa Panda robot with Omron mobile base.

```python theme={null}
EmbodimentTag.ROBOCASA_PANDA_OMRON
# Value: "robocasa_panda_omron"
```

#### GR1

The Fourier GR1 robot.

```python theme={null}
EmbodimentTag.GR1
# Value: "gr1"
```

### Posttrain embodiment tags

These embodiments are pre-registered for post-training:

#### UNITREE\_G1

The Unitree G1 robot.

```python theme={null}
EmbodimentTag.UNITREE_G1
# Value: "unitree_g1"
```

#### LIBERO\_PANDA

The Libero Panda robot.

```python theme={null}
EmbodimentTag.LIBERO_PANDA
# Value: "libero_panda"
```

#### OXE\_GOOGLE

The Open-X-Embodiment Google robot.

```python theme={null}
EmbodimentTag.OXE_GOOGLE
# Value: "oxe_google"
```

#### OXE\_WIDOWX

The Open-X-Embodiment WidowX robot.

```python theme={null}
EmbodimentTag.OXE_WIDOWX
# Value: "oxe_widowx"
```

#### OXE\_DROID

The Open-X-Embodiment DROID robot with relative joint position actions.

```python theme={null}
EmbodimentTag.OXE_DROID
# Value: "oxe_droid"
```

#### BEHAVIOR\_R1\_PRO

The Behavior R1 Pro robot.

```python theme={null}
EmbodimentTag.BEHAVIOR_R1_PRO
# Value: "behavior_r1_pro"
```

### Custom embodiments

#### NEW\_EMBODIMENT

Placeholder tag for any new embodiment during post-training.

```python theme={null}
EmbodimentTag.NEW_EMBODIMENT
# Value: "new_embodiment"
```

## Usage

```python theme={null}
from gr00t.data.embodiment_tags import EmbodimentTag
from gr00t.data.types import VLAStepData

# Create VLA step data with embodiment tag
vla_step = VLAStepData(
    images={"front_cam": [image_array]},
    states={"joint_positions": state_array},
    actions={"joint_velocities": action_array},
    text="Pick up the apple",
    embodiment=EmbodimentTag.UNITREE_G1,
)

# Access embodiment value
print(vla_step.embodiment.value)  # "unitree_g1"
```