Skip to content
Robotics

X-VLA Piper

18 чекпоинтов файнтюна для робота Piper

1 min read

X-VLA Piper от SXY7 — полнопараметрический файнтюн базовой модели 2toINF/X-VLA-Pt для задачи nesting_3dolls. В репозитории 18 полных чекпоинтов с шага 30000 по 200000: веса, конфиги, манифест и состояние обучения. Это VLA-модель для робототехники: по RGB-видам и положению суставов она предсказывает действия.

Ключевые особенности

  • 18 чекпоинтов: шаги 30000–200000, примерно по 10,56 ГБ каждый.
  • Вход: верхний RGB и RGB с правого запястья 224×224; третий вид замаскирован.
  • 7 измеренных значений суставов/гриппера как проприоцепция.
  • 30 предсказанных абсолютных 7-мерных действий суставов/гриппера.
  • В каждом чекпоинте model.safetensors, конфиги, piper_manifest.json и training_state.pt.
  • Нужно сохранять статистики нормализации и domain_id из манифеста.

Пример

Сценарий: подать два RGB-кадра 224×224 и 7 значений суставов, сохранить domain_id из piper_manifest.json и загрузить модель через XVLA и XVLAProcessor из совместимого исходника. На выходе — 30 абсолютных 7-мерных действий для робота Piper.

ПодробнееБаза: 2toINF/X-VLA-Pt, полнопараметрический файнтюн. Нужны upstream-коммит 6c4ecfb8be458ae7c0c96094e4b0add4174f4fe2, Transformers 4.51.3 и конфиг action_mode=auto, real_action_dim=7, max_action_dim=20.
Восстановительный тренер не сохранил точный курсор итерируемого датасета.
Каждый чекпоинт весит около 10,56 ГБ — учитывайте место на диске.
Discuss Telegram 0

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 3.3 GB · ready files4.1 GB
  • FP8 / INT8recommendedweights 1.6 GB2.3 GB
  • 4-bitweights 0.9 GB1.5 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.

    # SXY7/X-VLA-Piper-nesting_3dolls
    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 SXY7/X-VLA-Piper-nesting_3dolls --local-dir ~/models/X-VLA-Piper-nesting_3dolls
    # 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. Florence-2-largefamily root
  2. X-VLA-Pt
  3. X-VLA-Piper-nesting_3dolls

New posts about versions of Florence-2-large (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 3All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 10All on Hugging Face
Merges 1All on Hugging Face

About the model

SXY7/X-VLA-Piper-nesting_3dolls
Author
SXY7
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

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

Comments