Skip to content
Robotics

pi05 VLA для UR5

политика захвата CPU

1 min readupdated

Это Vision-Language-Action политика, файнтюн lerobot/pi05_base от Physical Intelligence, адаптированная в LeRobot. Модель на ~4,14 млрд параметров обучена для робота UR5 с одной камерой: захватить CPU вакуумным захватом и положить в коробку.

Ключевые особенности

  • Обучена на 500 эпизодах и 322 402 кадрах при 30 FPS.
  • Вход: изображение (3, 480, 640) и состояние (7,); выход: действие (7,).
  • 20 000 шагов обучения, batch size 32, AdamW, learning rate 2.5e-05.
  • Запуск и дообучение через CLI LeRobot; лицензия apache-2.0.

Пример

lerobot-rollout \
  --strategy.type=base \
  --robot.type=ur5 \
  --robot.port=<your_robot_port> \
  --robot.cameras="{ camera: {type: opencv, index_or_path: 0, width: 640, height: 480, fps: 30}}" \
  --policy.path=SamerAyyad/VLA_Samer_04.11.2026_500eps_policy_Pi05 \
  --task="Pick up the CPU with the suction gripper and place it in the box" \
  --duration=60

Замените <...> на свои значения. Имена камер должны совпадать с observation-ключами, на которых обучалась политика.

ПодробнееБенчмарков нет: в карточке не приведены результаты оценки на реальном роботе.
Данные: 500 эпизодов, 322 402 кадра, 30 FPS; задача — перенос CPU вакуумным захватом в коробку.
Обучение: 20 000 шагов, batch 32, AdamW, lr 2.5e-05, seed 1000, LeRobot 0.6.2.
Требования к VRAM/железу не указаны; для запуска нужны UR5, камера и порт робота.
Discuss Telegram 0 27

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 8.7 GB · ready files10 GB
  • FP8 / INT8recommendedweights 3.9 GB4.7 GB
  • 4-bitweights 2.2 GB2.9 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 2080 Ti

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.

    # SamerAyyad/VLA_Samer_04.11.2026_500eps_policy_Pi05
    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 SamerAyyad/VLA_Samer_04.11.2026_500eps_policy_Pi05 --local-dir ~/models/VLA_Samer_04.11.2026_500eps_policy_Pi05
    # 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. paligemma-3b-pt-224family root
  2. pi05_base
  3. VLA_Samer_04.11.2026_500eps_policy_Pi05

New posts about versions of paligemma-3b-pt-224 (quants, finetunes, LoRAs) will show up under the bell.

Fine-tunes 12All on Hugging Face
LoRAs and adapters 12All on Hugging Face

About the model

SamerAyyad/VLA_Samer_04.11.2026_500eps_policy_Pi05
Author
SamerAyyad
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

270

Ask AI about this model

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

Comments

Other

SANA-RWM

SANA-RWM