Skip to content
Robotics

smolvla_real_stack_red_blue_100

SmolVLA-политика для укладки кубиков

1 min readupdated

Файнтюн lerobot/smolvla_base от автора tsangb34 для роборуки SO-101. Модель на 0,45 млрд параметров учится по языковой инструкции, состоянию суставов и трём RGB-камерам складывать красный кубик на синий.

Ключевые особенности

  • Одна задача: «Put the red cube on top of the blue cube».
  • Входы: 6-D observation.state и три камеры 256×256.
  • Выход: 6-D действие чанками по 50.
  • Обучение: 100 000 шагов, batch size 32, lr 0.0001, seed 1000.
  • Финальный чекпоинт — шаг 100 000.

Пример

from lerobot.policies.smolvla.modeling_smolvla import SmolVLAPolicy

policy = SmolVLAPolicy.from_pretrained(
    "tsangb34/smolvla_real_stack_red_blue_100"
)
ПодробнееВ симуляции по 50 эпизодов на чекпоинт: 20k — 60%, 40k — 78%, 60k — 80%, 80k — 86%, 100k — 78%.
Данные: датасет so101_real_stack_red_blue_100, одна задача в MuJoCo so101-libero-mini.
Ограничение: оценка только в симуляции, без данных о VRAM и переносе на реальную руку.
Discuss Telegram 0 26

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.

    # tsangb34/smolvla_real_stack_red_blue_100
    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 tsangb34/smolvla_real_stack_red_blue_100 --local-dir ~/models/smolvla_real_stack_red_blue_100
    # 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_real_stack_red_blue_100

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

tsangb34/smolvla_real_stack_red_blue_100
Author
tsangb34
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

260

Ask AI about this model

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

Comments