You can deploy this model and start earning money today!

MoAIBo/
handover_scene1_2_giver_receiver_merged_pos_vel_policy_vision_expert

Name: handover_scene1_2_giver_receiver_merged_pos_vel_policy_vision_expert
Availability: InStock

MoAIBo/handover_scene1_2_giver_receiver_merged_pos_vel

Views

No views yet

Model Card for smolvla

SmolVLA is a compact, efficient vision-language-action model that achieves competitive performance at reduced computational costs and can be deployed on consumer-grade hardware.

This policy has been trained and pushed to the Hub using LeRobot.

Learn how to train and run it in the LeRobot smolvla guide, or browse the full documentation.

Model Details

License: apache-2.0
Fine-tuned from: lerobot/smolvla_base
Robot type: so101_tb4
Cameras: camera_left, camera_right, camera_wrist, camera_d455, depth

Inputs & Outputs

The policy consumes these observation features and produces these action features.

Inputs

Feature	Type	Shape
`observation.state`	STATE	`(11,)`
`observation.images.camera_left`	VISUAL	`(3, 360, 640)`
`observation.images.camera_right`	VISUAL	`(3, 360, 640)`
`observation.images.camera_wrist`	VISUAL	`(3, 360, 640)`
`observation.images.camera_d455`	VISUAL	`(3, 360, 640)`
`observation.images.depth`	VISUAL	`(3, 360, 640)`

Outputs

Feature	Type	Shape
`action`	ACTION	`(8,)`

Training Dataset

Repository: MoAIBo/handover_scene1_2_giver_receiver_merged_pos_vel
Episodes: 338
Frames: 354884
Frame rate: 30 FPS
Task(s): " Approach Robot 1, present and handover black spatula, release grasp once Robot 1 has a secure grasp on it, then go your separate way.", " Approach Robot 0, reach out and take the black spatula from Robot 0, and once you have a secure grasp on it go your separate way.", " Approach Robot 1, present and handover blue brush, release grasp once Robot 1 has a secure grasp on it, then go your separate way.", " Approach Robot 0, reach out and take the blue brush from Robot 0, and once you have a secure grasp on it go your separate way.", " Approach Robot 1, present and handover screwdriver, release grasp once Robot 1 has a secure grasp on it, then go your separate way.", " Approach Robot 0, reach out and take the screwdriver from Robot 0, and once you have a secure grasp on it go your separate way.", " Approach Robot 1, present and handover bottle, release grasp once Robot 1 has a secure grasp on it, then go your separate way.", " Approach Robot 0, reach out and take the bottle from Robot 0, and once you have a secure grasp on it go your separate way.", " Approach Robot 0, reach out and take the spatula from Robot 0, and once you have a secure grasp on it go your separate way.", " Approach Robot 1, present and handover spatula, release grasp once Robot 1 has a secure grasp on it, then go your separate way."

Training Configuration

Setting	Value
Training steps	50000
Batch size	13
Optimizer	adamw
Learning rate	0.0001
Seed	1000
LeRobot version	0.6.0

How to Get Started with the Model

New to LeRobot? These guides cover the full workflow:

Install LeRobot — set up the lerobot package.
Hardware setup — assemble, wire, and calibrate your robot and cameras.
Record data & train a policy — the end-to-end imitation-learning walkthrough.
CLI cheat-sheet — quick reference for the lerobot-* commands.

The short version to run and train this policy:

Run the policy on your robot

bash

lerobot-rollout \
  --strategy.type=base \
  --robot.type=so101_tb4 \
  --robot.port=<your_robot_port> \
  --robot.cameras="{ <camera_1>: {type: opencv, index_or_path: <index_or_path>, width: 640, height: 480, fps: 30}, <camera_2>: {type: opencv, index_or_path: <index_or_path>, width: 640, height: 480, fps: 30}}" \
  --policy.path=MoAIBo/handover_scene1_2_giver_receiver_merged_pos_vel_policy_vision_expert \
  --task="<Robot0><Giver> Approach Robot 1, present and handover black spatula, release grasp once Robot 1 has a secure grasp on it, then go your separate way." \
  --duration=60

Replace the remaining <...> placeholders with your own values: --robot.port and the camera names/indices are specific to your machine, and the camera names must match the observation keys this policy was trained on.

When --strategy.type=base is used the script doesn't record the episodes. Skipping duration will make the policy run indefinitely. For more information look at rollout documentation.

Train your own policy

This policy type is usually fine-tuned from the pretrained base model lerobot/smolvla_base:

bash

lerobot-train \
  --dataset.repo_id=${HF_USER}/<dataset> \
  --policy.path=lerobot/smolvla_base \
  --output_dir=outputs/train/<policy_repo_id> \
  --job_name=lerobot_training \
  --policy.device=cuda \
  --policy.repo_id=${HF_USER}/<policy_repo_id> \
  --wandb.enable=true

Writes checkpoints to outputs/train/<policy_repo_id>/checkpoints/.

Evaluation

No evaluation results have been provided for this policy yet.

Citation

If you use this policy, please cite the method linked in the description above, along with LeRobot:

bibtex

@misc{cadene2024lerobot,
    author = {Cadene, Remi and Alibert, Simon and Soare, Alexander and Gallouedec, Quentin and Zouitine, Adil and Palma, Steven and Kooijmans, Pepijn and Aractingi, Michel and Shukor, Mustafa and Aubakirova, Dana and Russi, Martino and Capuano, Francesco and Pascal, Caroline and Choghari, Jade and Moss, Jess and Wolf, Thomas},
    title = {LeRobot: State-of-the-art Machine Learning for Real-World Robotics in Pytorch},
    howpublished = "\url{https://github.com/huggingface/lerobot}",
    year = {2024}
}