Downloads · 30 days
11
24% of all-time downloads
dp_so_101_red_foam_ball
Yu-Zhou-Wang/dp_so_101_red_foam_ball
dp_so_101_red_foam_ball is a robotics model from Yu-Zhou-Wang. 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. The card lists the license as apache-2.0.
LeRobot Diffusion Policy trained on Jingyi-Z/sotac episodes 0–20 (Pick up the red foam ball and place it into the container).
All-time downloads
45
Public
Parameters
278M
1.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.1 GB · 100%
At a glance
- Task
- Robotics
- Library
- lerobot
- License
- apache-2.0
- Access
- Public
- Created
- Sep 2, 2026
- Updated
- Sep 2, 2026
- SHA
- 6a55c0db
Model card
From the Hugging Face model README
Diffusion Policy — SO-101 red foam ball
LeRobot Diffusion Policy trained on Jingyi-Z/sotac episodes 0–20 (Pick up the red foam ball and place it into the container).
Trained from scratch (ResNet18 ImageNet backbone, one encoder per camera). Tactile is not used.
| Robot | SO-101 follower, 6 joint positions |
| Cameras | observation.images.top, observation.images.wrist (640×480) |
| Horizon / n_obs / n_action | 64 / 2 / 32 |
| Steps | 100000 (this checkpoint) |
| Batch | 8 |
Load
from lerobot.policies.diffusion.modeling_diffusion import DiffusionPolicy
policy = DiffusionPolicy.from_pretrained("Yu-Zhou-Wang/dp_so_101_red_foam_ball")
Real-robot eval (LeRobot 0.6). Camera keys must be top and wrist (or rename front → top):
lerobot-record \
--robot.type=so101_follower \
--policy.path=Yu-Zhou-Wang/dp_so_101_red_foam_ball \
--dataset.repo_id=<you>/eval_dp_red_foam_ball
Training code lives on Delta at dp_so_101. Dataset: Jingyi-Z/sotac.
- Task
- Robotics
- Library
- lerobot
- License
- apache-2.0
- Created
- Sep 2, 2026
- Updated
- Sep 2, 2026