CoRL2026-CSI/Isaaclab-so101_11task_baseCaP_3300epi_10fps
Viewer • Updated • 1.18M • 28
How to use CoRL2026-CSI/smolVLA-IsaacLab-Multi-Task-8epoch-mod with LeRobot:
# See https://github.com/huggingface/lerobot?tab=readme-ov-file#installation for more details git clone https://github.com/huggingface/lerobot.git cd lerobot pip install -e .[smolvla]
# Launch finetuning on your dataset python lerobot/scripts/train.py \ --policy.path=CoRL2026-CSI/smolVLA-IsaacLab-Multi-Task-8epoch-mod \ --dataset.repo_id=lerobot/svla_so101_pickplace \ --batch_size=64 \ --steps=20000 \ --output_dir=outputs/train/my_smolvla \ --job_name=my_smolvla_training \ --policy.device=cuda \ --wandb.enable=true
# Run the policy using the record function
python -m lerobot.record \
--robot.type=so101_follower \
--robot.port=/dev/ttyACM0 \ # <- Use your port
--robot.id=my_blue_follower_arm \ # <- Use your robot id
--robot.cameras="{ front: {type: opencv, index_or_path: 8, width: 640, height: 480, fps: 30}}" \ # <- Use your cameras
--dataset.single_task="Grasp a lego block and put it in the bin." \ # <- Use the same task description you used in your dataset recording
--dataset.repo_id=HF_USER/dataset_name \ # <- This will be the dataset name on HF Hub
--dataset.episode_time_s=50 \
--dataset.num_episodes=10 \
--policy.path=CoRL2026-CSI/smolVLA-IsaacLab-Multi-Task-8epoch-modlerobot/smolvla_base 를 IsaacLab 시뮬레이션 SO101 11-task 멀티태스크 데이터셋 CoRL2026-CSI/Isaaclab-so101_11task_baseCaP_3300epi_10fps 으로 8 epoch 파인튜닝한 SmolVLA 정책.
이 체크포인트는 full model (model.safetensors) 입니다 — LoRA adapter 가 아니며, 그대로 로드해 사용합니다.
lerobot/smolvla_base (SmolVLM2-500M-Video-Instruct VLM + action expert)top, left_wrist (480×640) — 정책 키 camera1(left_wrist) / camera2(top) 로 renameobservation.state[6] + 카메라 2개 + language instruction (task)action[6] (joint position)chunk_size=50, n_action_steps=50VLM frozen + action expert only — SmolVLA 공식 표준 학습 방식 (SmolVLA paper, arXiv:2506.01844).
| 구성요소 | 상태 |
|---|---|
| VLM backbone (SmolVLM2) | ❄️ 완전 Frozen (freeze_vision_encoder=true) |
| Action expert | 🔥 학습 (train_expert_only=true) |
| PEFT / LoRA | 사용 안 함 |
| 항목 | 값 |
|---|---|
| Dataset | Isaaclab-so101_11task_baseCaP_3300epi_10fps — 3,300 episodes / 1,175,352 frames / 11 tasks / 10 fps |
| Epochs / Steps | 8 epoch / 36,800 steps |
| Global batch size | 256 (micro batch 128 × 2 GPU) |
| Optimizer | AdamW — lr 1e-4, weight_decay 1e-10, grad_clip_norm 10.0 |
| LR scheduler | cosine_decay_with_warmup — warmup 1,000 / decay 30,000 / peak_lr 1e-4 / decay_lr 2.5e-6 |
| chunk_size / n_action_steps | 50 / 50 |
| Seed | 1000 |
| Dataloader workers | 16 |
| Mixed precision | no (bf16 inference) |
| Image augmentation | ColorJitter (brightness/contrast/saturation/hue) + SharpnessJitter — 기하학적 변형(회전/이동/반전) 없음 (VLA 좌우 의미 보존) |
| Hardware | 2 × NVIDIA H100 80GB |
| Final loss | 0.020 |
LeRobot dataset 의 카메라 키와 SmolVLA 정책 키 매핑:
| Dataset key | Policy key |
|---|---|
observation.images.left_wrist |
observation.images.camera1 |
observation.images.top |
observation.images.camera2 |
추론·평가 시 반드시 위와 동일한 rename 을 적용해야 합니다 (학습-추론 일관성).
observation.state[6] (joint position) + 카메라 2개 + language instruction(task) 만action[6] (joint position) 만ee_pos / gripper_binary / state.radian_urdf0 / action.radian_urdf0 는 학습에서 제외camera1/2/3)로 고정이라 camera3 슬롯이 config 에 존재하지만, 데이터셋 카메라는 2개뿐이라 실제로 데이터가 흐르는 카메라는 2개입니다.from lerobot.policies.smolvla.modeling_smolvla import SmolVLAPolicy
policy = SmolVLAPolicy.from_pretrained("CoRL2026-CSI/smolVLA-IsaacLab-Multi-Task-8epoch-mod")
Built on top of LeRobot and the SmolVLA base checkpoint. Project: CoRL 2026 CSI submission.
Base model
lerobot/smolvla_base