You can deploy this model and start earning money today!

Faless/
harvest_apples_smolvla_v2_action_expert_bs96_ee_20phrases

Name: harvest_apples_smolvla_v2_action_expert_bs96_ee_20phrases
Availability: InStock

Faless/harvest_apples_with_agilex_piper_sim_ee_paraphrases20

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: piper_full
Cameras: ego, front

Inputs & Outputs

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

Inputs

Feature	Type	Shape
`observation.state`	STATE	`(8,)`
`observation.images.camera1`	VISUAL	`(3, 256, 256)`
`observation.images.camera2`	VISUAL	`(3, 256, 256)`

Outputs

Feature	Type	Shape
`action`	ACTION	`(7,)`

Training Dataset

Repository: Faless/harvest_apples_with_agilex_piper_sim_ee_paraphrases20
Episodes: 600
Frames: 1020000
Frame rate: 25 FPS
Task(s): "Collect each red apple separately and put it into the green basket.", "Collect the red apples one by one and put them in the green basket.", "Gather the red apples individually, placing each one in the green basket.", "Grab each red apple separately and place it in the green basket.", "Harvest the red apples one by one and deposit them in the green basket.", "Move each red apple, one at a time, into the green basket.", "Move the red apples to the green basket one after another.", "One by one, pick the red apples and place them in the green basket.", "Pick and place the red apples into the green basket individually.", "Pick each red apple and place it into the green basket.", "Pick the red apples one by one and place them into the green basket.", "Pick up one red apple at a time and put it into the green basket.", "Pick up the red apples individually and transfer them to the green basket.", "Place every red apple in the green basket, handling one apple at a time.", "Place the red apples into the green basket one at a time.", "Put every red apple into the green basket, one at a time.", "Sequentially pick the red apples and put each one into the green basket.", "Take one red apple at a time and place it inside the green basket.", "Take the red apples individually and place each one in the green basket.", "Transfer the red apples individually to the green basket."

Training Configuration

Setting	Value
Training steps	55000
Batch size	96
Optimizer	adamw
Learning rate	0.0001
Seed	1000
LeRobot version	0.5.2

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=piper_full \
  --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=Faless/harvest_apples_smolvla_v2_action_expert_bs96_ee_20phrases \
  --task="Collect each red apple separately and put it into the green basket." \
  --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}
}