Downloads · 30 days
6
17% of all-time downloads
so100-smolvla-new60-ft-s2
tonghuiwang123/so100-smolvla-new60-ft-s2
so100-smolvla-new60-ft-s2 is a robotics model from tonghuiwang123. Use it for the robotics task on the model card, and read the license before you ship it in a product. It is set up for lerobot.
SMOLVLA policy for SO-ARM100 / SO100 — Grab the white cube to the white cup
All-time downloads
35
Public
Parameters
450M
15.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors907 MB · 100%
Parameter types
How the weights are stored.
BF16447M · 99%
Model card
From the Hugging Face model README
so100-smolvla-new60-ft-s2 · step 85000
SMOLVLA policy for SO-ARM100 / SO100 — Grab the white cube to the white cup
- 数据:
tonghuiwang123/4060 episodes / 13454 frames / 2×1280x720 @30fps - Checkpoint: step 85000 ≈ 404.3 epoch（有效 batch 64）
- 两阶段 seed=2000；与 so100-smolvla-new60-ft (seed=1000) 构成种子方差对照，其余配置完全相同
相机命名（各算法族不同，顺序错了不报错但表现会莫名变差）
| 物理位置 | 设备 | 本模型要求的 key |
|---|---|---|
| 顶部 (top-down) | /dev/video2 | camera1 |
| 腕部 (wrist) | /dev/video4 | camera2 |
部署
lerobot-record \
--robot.type=so100_follower --robot.port=/dev/ttyACM0 \
--robot.id=my_awesome_follower_arm \
--robot.cameras="{
camera1: {type: opencv, index_or_path: /dev/video2, width: 1280, height: 720, fps: 30},
camera2: {type: opencv, index_or_path: /dev/video4, width: 1280, height: 720, fps: 30},
}" \
--dataset.repo_id=tonghuiwang123/eval_so100-smolvla-new60-ft-s2 \
--dataset.single_task="Grab the white cube to the white cup" \
--policy.path=tonghuiwang123/so100-smolvla-new60-ft-s2
分辨率必须 1280x720。推理显存 1.0 GB。
中间 checkpoint 用 --policy.pretrained_revision=ckpt-XXXXXX。
- Task
- Robotics
- Library
- lerobot
- Created
- Aug 29, 2026
- Updated
- Aug 29, 2026