EN

LingBot-VA base

LingBot-VA base

Checkpoint

Autoregressive model that jointly predicts future video and robot actions.

Default checkpoint

lerobot/lingbot_va_base

openApache License 2.0

Other releases

1

published checkpoints

Supported robots

No verified robot compatibility recorded.

Benchmark results

No published benchmark results.

Specifications

Architecture components: Future video generation
Backbone: Autoregressive video-action transformer
Frameworks: LeRobot / PyTorch
Input modalities: Language instruction
Input modalities: RGB camera images
Input modalities: Robot state / proprioception
Languages: English
Model type: Vision-language-action model
Model type: World-action model
Output action space: Future video chunks and robot action chunks
Parameter count: 5.0889B parameters
Pipeline task: Robotics

Lineage

Unknown

Architecture & I/O

trajectory · continuous

Unknown

Future video chunks and robot action chunks

Source · lerobot

Deployment

LeRobot / PyTorch · NVIDIA GPU

bfloat16 · 24 GB VRAM · unknown

Source · lerobot

Resources

documentation · LeRobot lingbot_va documentation weights · lerobot/lingbot_va_base model card

Submit a source or correction