Skip to content
Robotics

X-VLA для UR5e + Schunk

политика захвата и переноса блока

2 min read

Это policy от Leberkaesweckle для реального робота UR5e с гриппером Schunk, обученная в LeRobot на базе lerobot/xvla-base. X-VLA — vision-language-action фреймворк с soft prompt эмбеддингами: разные роботы и сенсоры кодируются как отдельные «задачи». Модель на ~0.88 млрд параметров принимает несколько изображений и 8-мерное состояние, а выдаёт 7-мерное действие.

Ключевые особенности

  • Дообучена под робота ur5_schunk_follower; камеры front и side.
  • Вход: несколько VISUAL-тензоров (256×256 и 224×224) и observation.state размером (8,).
  • Выход: action размером (7,).
  • Обучение на 50 эпизодах, 15 011 кадрах при 15 FPS; задача — схватить оранжевый блок и опустить его в синюю коробку.
  • 20 000 шагов, batch size 32, learning rate 0.0001, оптимизатор xvla-adamw, LeRobot 0.6.2.
  • Лицензия apache-2.0.

Пример запуска

Ключевые параметры; порт и индексы камер подставьте свои.

lerobot-rollout \
  --strategy.type=base \
  --robot.type=ur5_schunk_follower \
  --robot.port=<your_robot_port> \
  --robot.cameras="{ front: {type: opencv, index_or_path: <idx1>, width: 640, height: 480, fps: 30}, side: {type: opencv, index_or_path: <idx2>, width: 640, height: 480, fps: 30}}" \
  --policy.path=Leberkaesweckle/Dataset_UR5e_Schunk_Gripper_50_episodes_10_09_2026_XVLA \
  --task="Grip the orange block and drop them into the blue box" \
  --duration=60
Подробнее Публичных бенчмарков нет: в карточке «No evaluation results».
Требования к VRAM не указаны; в примере обучения используется --policy.device=cuda.
Порт робота и индексы камер нужно подставить свои; имена камер должны совпадать с observation keys.
Датасет: 50 эпизодов, 15 011 кадров, 15 FPS; обучение — 20 000 шагов.
Discuss Telegram 0 25

Will it run?

Working out the memory needs

The bot is still collecting this model's weight sizes; the estimate will appear here in a few minutes. Meanwhile you can rent a GPU with headroom.

Massed ComputeOpen Massed Compute

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # Leberkaesweckle/Dataset_UR5e_Schunk_Gripper_50_episodes_10_09_2026_XVLA
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download Leberkaesweckle/Dataset_UR5e_Schunk_Gripper_50_episodes_10_09_2026_XVLA --local-dir ~/models/Dataset_UR5e_Schunk_Gripper_50_episodes_10_09_2026_XVLA
    # The model is in ~/models. How to run it is in the model card (link above).
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of Dataset_UR5e_Schunk_Gripper_50_episodes_10_09_2026_XVLA (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

Leberkaesweckle/Dataset_UR5e_Schunk_Gripper_50_episodes_10_09_2026_XVLA
Author
Leberkaesweckle
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

250

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments

Robotics

GR00T G1

GR00T G1