---
license: apache-2.0
base_model: robbyant/lingbot-vla-v2-6b
tags: [robotics, vla, lingbot-vla-2, maniguard]
---

# LingBot-VLA 2.0 — ManiGuard datagen-v1 `lid` (joint, 2-cam)

LingBot-VLA 2.0 post-trained on the ManiGuard `lid` family, one of five base models
evaluated on ManiGuard-Bench under identical data, cameras, and controller.

- **Warm start:** [`robbyant/lingbot-vla-v2-6b`]( https://huggingface.co/robbyant/lingbot-vla-v2-6b )
  — the **pretrain** release (not the RoboTwin post-trained variant).
- **Data:** `IDEAS-Lab-Northwestern/datagen-lid-v1-joint-5cam` (LeRobot v2.1, consumed directly — no format conversion).
- **Inputs:** 2 cameras (`camera_top` = the `left` overview, `camera_wrist_left` = wrist),
  8-D joint state (7 arm + gripper) mapped into the 55-D unified vector.
- **Actions:** **absolute** joint targets (`subtract_state: false` on both features, per
  LingBot's own simulation recipe) — apply directly to a JointController, no delta step.
- **Recipe:** upstream post-training config unchanged (MoE + depth/DINO distillation on);
  global batch 256 = micro 32 x 8 GPUs, lr 5e-5 cosine, **2 epochs** (8,250 steps).
- **Ladder:** checkpoints every 2,063 steps.

Serving needs `maniguard/norm_stats.json` + `maniguard/robot_config.yaml`, both included here.