VibeCuisine/cucumber_grasp_act_train_vibepi_20260922-v1-act

<!-- vibedata:model-card -->

cucumbergraspacttrainvibepi_20260922-v1-act

act policy trained via Vibe Data Studio on 20 teleop episodes.

At a glance

Tasks

Language-conditioned — pass one of these instructions at inference:

Rig

• —Action/state (7-dim): shoulderpan, shoulderlift, elbowflex, wristflex, wrist_roll, gripper, tilt
• —Cameras @ 20 fps: base 640×480, top 640×480, wrist 640×480
• —Consumed image inputs (the checkpoint's resolved VISUAL features, from its config.json): observation.images.base, observation.images.top, observation.images.wrist

Data provenance

Materialized collection cucumber_grasp_act_train_vibepi_20260922@v1 (frozen 2026-09-23):

• —autoresearch-vibepi-000655-task-cucumber_grasp @ d55dfe57 (1 eps)
• —autoresearch-vibepi-000656-task-cucumber_grasp @ ffba74e6 (1 eps)
• —autoresearch-vibepi-000658-task-cucumber_grasp @ da404a1a (1 eps)
• —autoresearch-vibepi-000659-task-cucumber_grasp @ b6d772e0 (1 eps)
• —autoresearch-vibepi-000662-task-cucumber_grasp @ 96c4abd1 (1 eps)
• —autoresearch-vibepi-000663-task-cucumber_grasp @ a1811afc (1 eps)
• —autoresearch-vibepi-000664-task-cucumber_grasp @ c28024a8 (1 eps)
• —autoresearch-vibepi-000665-task-cucumber_grasp @ a1d7f8c5 (1 eps)
• —autoresearch-vibepi-000666-task-cucumber_grasp @ 54a5542c (1 eps)
• —autoresearch-vibepi-000667-task-cucumber_grasp @ d94112d5 (1 eps)
• —autoresearch-vibepi-000669-task-cucumber_grasp @ f6f28917 (1 eps)
• —autoresearch-vibepi-000670-task-cucumber_grasp @ faa2dca4 (1 eps)
• —autoresearch-vibepi-000672-task-cucumber_grasp @ 6700dd0e (1 eps)
• —autoresearch-vibepi-000673-task-cucumber_grasp @ 125ac09a (1 eps)
• —autoresearch-vibepi-000674-task-cucumber_grasp @ 1b540961 (1 eps)
• —autoresearch-vibepi-000676-task-cucumber_grasp @ c030195d (1 eps)
• —autoresearch-vibepi-000677-task-cucumber_grasp @ c7919c45 (1 eps)
• —autoresearch-vibepi-000679-task-cucumber_grasp @ 79719b97 (1 eps)
• —autoresearch-vibepi-000680-task-cucumber_grasp @ 5ff078c6 (1 eps)
• —autoresearch-vibepi-000681-task-cucumber_grasp @ da74aa34 (1 eps)

<details> <summary><b>Reproduce</b> — exact training command</summary>

lerobot-train --policy.type=act --policy.repo_id=VibeCuisine/cucumber_grasp_act_train_vibepi_20260922-v1-act --dataset.repo_id=VibeCuisine/cucumber_grasp_act_train_vibepi_20260922-v1-flat --dataset.revision=df4cdc9d6da4fde0467ab3b2646a31f8e9aee937 --batch_size=8 --steps=80000 --output_dir=outputs/train/cucumber_grasp_act_train_vibepi_20260922-v1-act --job_name=cucumber_grasp_act_train_vibepi_20260922-v1-act --policy.device=cuda --wandb.enable=true --wandb.disable_artifact=true --wandb.project=lerobot --wandb.run_id=vds64-9d2fc448 --wandb.entity=jeremyhx-freelance

--num_workers is appended host-side (clamped to the host's allocation).

</details>

Card generated by Vibe Data Studio on run completion.