Skip to content
Robotics

act_franka_sweep_v3

Robot Control Policy for Sweeping Task

2 min readupdated

This model is a ready-to-use control policy for a robotic manipulator, trained using ACT (Action Chunking with Transformers — imitation learning method that predicts short sequences of actions instead of individual steps). Designed for the Franka Panda robot, it handles sweeping an orange cube using visual camera data and joint states. It was trained using the LeRobot library on teleoperation demonstration data and enables autonomous task execution on real hardware.

Key features

  • Compatibility with the kinetic-type Franka Panda robot.
  • Use of dual visual streams from the wrist camera (regular and depth).
  • Training on a dataset of 49 episodes (6,842 frames at 10 FPS).
  • Built on the ACT architecture using the LeRobot framework version 0.6.2.

What you can do

  • Run the ready-to-use policy on a real Franka Panda robot to perform the cube-sweeping task.
  • Fine-tune or use the model as a baseline for your own robotic imitation learning tasks.

Example

Running the policy on your robot using a terminal command:

lerobot-rollout \
  --strategy.type=base \
  --robot.type=franka_panda_kinesthetic \
  --robot.port= \
  --robot.cameras="{ : {type: opencv, index_or_path: , width: 640, height: 480, fps: 30}, : {type: opencv, index_or_path: , width: 640, height: 480, fps: 30}}" \
  --policy.path=Matghost7/act_franka_sweep_v3 \
  --task="Sweep the orange cube" \
  --duration=60
Read moreDetails
Dataset: Matghost7/franka_sweep_cube_v3_merged
Training parameters: 50,000 steps, batch size 8, adamw optimizer, learning rate 1e-05
Input data: state (16,) and wrist images 3x480x640
Output data: actions of dimensionality (16,)
Discuss Telegram 0 23

Will it run?

Runs in F32

Fits without quantisation.

  • F32weights 0.2 GB · ready files0.7 GB
  • BF16weights 0.1 GB0.6 GB
  • FP8 / INT8recommendedweights 0.0 GB0.6 GB
  • 4-bitweights 0.0 GB0.5 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 · F32 — 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.

    # Matghost7/act_franka_sweep_v3
    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 Matghost7/act_franka_sweep_v3 --local-dir ~/models/act_franka_sweep_v3
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to act_franka_sweep_v3
  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 act_franka_sweep_v3 (quants, finetunes, LoRAs) will show up under the bell.

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

About the model

Matghost7/act_franka_sweep_v3
Author
Matghost7
Category
Robotics
Released
Posted

On NeyroLearn

Hugging Face

230

Ask AI about this model

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

Comments