Downloads · 30 days
5
8% of all-time downloads
act_dual_arm_m1
mateuszwasko1/act_dual_arm_m1
act_dual_arm_m1 is a robotics model from mateuszwasko1. 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.
This model is a baseline Action Chunking with Transformers (ACT) policy trained for cooperative dual-arm manipulation in PyBullet. It was trained without domain randomization to establish a performance baseline on a s…
All-time downloads
60
Public
Parameters
51.6M
206 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors206 MB · 100%
At a glance
- Task
- Robotics
- Library
- lerobot
- Access
- Public
- Created
- Jun 26, 2026
- Updated
- Jun 26, 2026
- SHA
- 5fb49e54
Model card
From the Hugging Face model README
ACT Dual-Arm Vanilla (Baseline)
This model is a baseline Action Chunking with Transformers (ACT) policy trained for cooperative dual-arm manipulation in PyBullet. It was trained without domain randomization to establish a performance baseline on a single object (cracker_box).
Model Details
- Architecture: Action Chunking with Transformers (ACT) with a ResNet18 vision backbone.
- Task: Dual-arm cooperative lifting and placing of an object.
- Action Space: 12-D absolute joint angles in radians (6 degrees of freedom per arm, gripper state fixed).
- Vision: 3 camera streams (overhead, left wrist, right wrist) at 224x224 resolution.
- Training Data: 149 expert demonstrations collected at 20 FPS.
Performance
Achieved a 95% success rate and 100% grip rate on the in-distribution evaluation suite. The model exhibits highly deterministic chunk execution and low positional jitter compared to diffusion-based alternatives.
- Task
- Robotics
- Library
- lerobot
- Created
- Jun 26, 2026
- Updated
- Jun 26, 2026