Skip to content
Robotics

act_grasp_lever

ACT-политика для захвата рычага

1 min read

CGdigipen/act_grasp_lever — обученная в LeRobot политика на основе Action Chunking with Transformers (ACT), около 0.05 млрд параметров. Это imitation-learning метод, который предсказывает короткие чанки действий вместо одиночных шагов. Политика создана для робота so_follower с камерами RIGHT и HEAD и задачи «identify and grasp lever handle».

Ключевые особенности

  • ACT предсказывает короткие последовательности действий, а не один шаг за раз.
  • На входе — состояние робота (6,) и два изображения 480×640 с камер RIGHT и HEAD.
  • На выходе — вектор действия (6,); целевой робот — so_follower.
  • Обучение на телеоперационных данных: 115 эпизодов, 47 618 кадров при 30 FPS.
  • Чекпоинт для LeRobot 0.6.1, лицензия Apache-2.0; запуск через lerobot-rollout.

Пример

lerobot-rollout \
  --strategy.type=base \
  --robot.type=so_follower \
  --robot.port=<your_robot_port> \
  --policy.path=CGdigipen/act_grasp_lever \
  --task="identify and grasp lever handle" \
  --duration=60
ПодробнееОбучение: 100 000 шагов, batch size 8, AdamW, learning rate 1e-5, seed 1000, LeRobot 0.6.1.
Датасет: 115 эпизодов, 47 618 кадров, 30 FPS; задача «identify and grasp lever handle».
Оценки на реальном роботе в карточке нет — success rate не указан.
Политика привязана к so_follower и ключам камер RIGHT/HEAD; для другого железа нужно переобучение или свои порты/камеры.
Discuss Telegram 0 41

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.

    # CGdigipen/act_grasp_lever
    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 CGdigipen/act_grasp_lever --local-dir ~/models/act_grasp_lever
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to act_grasp_lever
  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 act_grasp_lever (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

CGdigipen/act_grasp_lever
Author
CGdigipen
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

410

Ask AI about this model

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

Comments

Robotics

gaze_act

gaze_act