Skip to content
Robotics

dobei/smolvla_so101_pickplace_v2

robotics fine-tune

2 min read

This model is a specialized control policy for an so_follower-type robotic arm, trained for pick-and-place tasks. It is based on the compact SmolVLA model (an efficient Vision-Language-Action architecture) running through the LeRobot library. The model was trained on demonstrations using three surveillance cameras and the manipulator's system state. This solution will be useful for researchers and engineers working on robotics using affordable hardware.

Key features

  • SmolVLA base: a compact and efficient open-source vision-language-action model.
  • Multiple cameras: support for three primary visual streams and additional state data.
  • LeRobot integration: full compatibility with the Hugging Face ecosystem for robot control.
  • Specific task: trained on object pick-and-place demonstrations.

What you can do

  • Run on a robotic arm: apply the ready-to-use policy to automate the pick-and-place process on compatible hardware.
  • Fine-tune the model: use the collected dataset and base model to customize it for your own manipulator tasks.

Example

lerobot-rollout \
  --strategy.type=base \
  --robot.type=so_follower \
  --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=dobei/smolvla_so101_pickplace_v2 \
  --task="pick and place" \
  --duration=60
Read moreDetails
License: apache-2.0
Dataset: dobe/so101_red_cube_blue_box_2_20261117_161737 (25 episodes, 11,286 frames at 30 FPS)
Training parameters: 10 steps, batch size 16, adamw optimizer, learning rate 0.0001
Discuss Telegram 0

Will it run?

Runs in F32

Fits without quantisation.

  • F32weights 1.7 GB · ready files2.3 GB
  • BF16weights 0.8 GB1.4 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 · 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.

    # dobei/smolvla_so101_pickplace_v2
    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 dobei/smolvla_so101_pickplace_v2 --local-dir ~/models/smolvla_so101_pickplace_v2
    # 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_so101_pickplace_v2

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

dobei/smolvla_so101_pickplace_v2
Author
dobei
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