Skip to content
Robotics

ACT SO-101

политика захвата кубика для LeRobot

2 min read

Это ACT-политика (Action Chunking with Transformers) для робота so_follower, обученная и опубликованная через LeRobot. У модели около 0,05B параметров (≈50M): на входе 6-мерное состояние и два изображения 240×320 с камер top и wrist, на выходе 6-мерное действие. Задача — взять красный кубик и положить его в серую коробку.

Ключевые особенности

  • ACT предсказывает короткие чанки действий вместо одиночных шагов и учится на телеоперационных данных.
  • Две камеры — top и wrist; наблюдения включают состояние (6,) и два VISUAL-входа (3, 240, 320).
  • Политика рассчитана на робота so_follower и запускается через lerobot-rollout.
  • Обучение: 5000 шагов, batch size 8, оптимизатор AdamW, learning rate 1e-05, seed 1000, LeRobot 0.6.2.
  • Датасет: 5 эпизодов, 2119 кадров, 30 FPS; задача — «Pick up the red cube and place it in the grey box.»
  • Результатов оценки на реальном роботе в карточке нет.

Пример

lerobot-rollout \
  --strategy.type=base \
  --robot.type=so_follower \
  --robot.port=<your_robot_port> \
  --policy.path=SebasKoedam/act_so101_training_2 \
  --task="Pick up the red cube and place it in the grey box." \
  --duration=60
ПодробнееБенчмарков нет: в карточке указано «No evaluation results».
Данные: 5 эпизодов и 2119 кадров при 30 FPS.
Обучение: 5000 шагов, batch size 8, AdamW, learning rate 1e-05, seed 1000, LeRobot 0.6.2.
Ограничение: узкая задача и мало данных; для запуска нужны порт робота и камеры, совпадающие с observation keys.
Discuss Telegram 0 34

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.

    # SebasKoedam/act_so101_training_2
    # ===== Tokens: paste yours between the quotes =====
    export HF_TOKEN=""        # huggingface.co/settings/tokens, for gated models
    # ==========================================
    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 SebasKoedam/act_so101_training_2 --local-dir ~/models/act_so101_training_2
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to act_so101_training_2
  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_so101_training_2 (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

SebasKoedam/act_so101_training_2
Author
SebasKoedam
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

340

Ask AI about this model

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

Comments