Skip to content
Other

Harness VLA pi0.5 Real - robotics model with text memory

2 min readupdated

A checkpoint of the VLA model (Vision-Language-Action, combining vision, language, and action control) based on the pi0.5 architecture, trained on real robot data. The model features a text memory module (scheme C) and is designed for deployment and inference on robotic hardware. Training took 10,000 optimizer steps, and the checkpoint successfully passed validation upon reloading. Suitable for robotics researchers and engineers working with physical devices.

Key features

  • Based on the pi0.5 architecture (Vision-Language-Action model)
  • Added text memory (scheme C) for sequential task execution
  • Trained on real robotics data totaling over 48 thousand rows
  • Full set of inference parameters weighs about 13.4 GB
  • Includes built-in weights and SAM 3-based tracker code for deployment

What you can do

  • Run model inference using the included policy.py script
  • Use the built-in SAM 3-based bowl tracker for grasping and manipulation tasks
  • Set up a deployment cycle for a real robot using the real_deploy.py script
  • Integrate configuration files and normalization statistics for integrity checking

Example

Example code for loading and running the model via policy.py can be found in the model repository under the source/ directory.

Read moreDetails
Optimizer steps: 10,000 (global batch 128)
Training data rows: 48,609
Final loss: 0.0041
Limitations: The model has not yet been tested on a real robot
Parameter weight: ~13.4 GB
Discuss Telegram 0

Will it run?

Runs in BF16

Fits without quantisation.

  • BF16weights 3.2 GB · ready files4.0 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.

    # HITzyx/harness-vla-pi05-real
    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 HITzyx/harness-vla-pi05-real --local-dir ~/models/harness-vla-pi05-real
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to harness-vla-pi05-real
  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 harness-vla-pi05-real (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

HITzyx/harness-vla-pi05-real
Author
HITzyx
Category
Other
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