Skip to content
Robotics

SmolVLA Sponge Pick

компактная VLA-политика для робота-манипулятора

1 min readupdated

Это дообученная политика SmolVLA от mharbovskyi: компактная vision-language-action модель на ~0,45 млрд параметров, которая управляет роботом so_follower по изображениям с трёх камер. Она обучена класть красную или жёлтую губку в миску. База — lerobot/smolvla_base, обучение и запуск — через LeRobot.

Ключевые особенности

  • Основана на lerobot/smolvla_base — компактной VLA-модели.
  • Дообучена под робота so_follower с камерами wrist, top и side.
  • Выполняет две задачи: «Put the red sponge in the bowl» и «Put the yellow sponge in the bowl».
  • Входы: состояние (6,) и три изображения 256×256; выход — действие (6,).
  • Запуск через lerobot-rollout, обучение через lerobot-train.
  • Датасет: 66 эпизодов, 27 675 кадров при 30 FPS; обучение — 20 000 шагов.

Пример

lerobot-rollout --strategy.type=base --robot.type=so_follower --robot.port=<your_robot_port> --policy.path=mharbovskyi/smolvla_sponge_pick_r_y_v1_1110_2339 --task="Put the red sponge in the bowl" --duration=60
ПодробнееОценки на реальном роботе в карточке нет — таблица успешности не заполнена.
Обучение: 20 000 шагов, batch size 32, AdamW, learning rate 1e-4, seed 1000, LeRobot 0.7.1.
Ограничения: политика заточена под so_follower и три камеры; датасет — 66 эпизодов, 27 675 кадров при 30 FPS.
Лицензия — Apache-2.0.
Discuss Telegram 0 14

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 0.8 GB · ready files1.4 GB
  • FP8 / INT8recommendedweights 0.4 GB1.0 GB
  • 4-bitweights 0.2 GB0.8 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 2060 · BF16 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

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.

    # mharbovskyi/smolvla_sponge_pick_r_y_v1_1110_2339
    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 mharbovskyi/smolvla_sponge_pick_r_y_v1_1110_2339 --local-dir ~/models/smolvla_sponge_pick_r_y_v1_1110_2339
    # 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

  1. SmolLM2-360Mfamily root
  2. SmolLM2-360M-Instruct
  3. SmolVLM-500M-Instruct
  4. SmolVLM2-500M-Video-Instruct
  5. smolvla_base
  6. smolvla_sponge_pick_r_y_v1_1110_2339

New posts about versions of SmolLM2-360M (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 8All on Hugging Face
Merges 1All on Hugging Face

About the model

mharbovskyi/smolvla_sponge_pick_r_y_v1_1110_2339
Author
mharbovskyi
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

140

Ask AI about this model

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

Comments