Downloads · 30 days
9
43% of all-time downloads
franka-30demos-randomized_task2_cartesian_20260921T003023Z-step-8000
NaaaaaiVe6/franka-30demos-randomized_task2_cartesian_20260921T003023Z-step-8000
franka-30demos-randomized_task2_cartesian_20260921T003023Z-step-8000 is a robotics model from NaaaaaiVe6. 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 openpi.
Action representation: absolute Cartesian XYZ + quaternion xyzw + binary gripper (-1/+1). These are not controller delta actions.
All-time downloads
21
Public
Parameters
3.6B
7.2 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors7.2 GB · 100%
At a glance
- Task
- Robotics
- Library
- openpi
- Access
- Public
- Created
- Sep 24, 2026
- Updated
- Sep 24, 2026
- SHA
- e1a7abee
Model card
From the Hugging Face model README
Franka Cartesian checkpoint, step 8000
Action representation: absolute Cartesian XYZ + quaternion xyzw + binary gripper (-1/+1). These are not controller delta actions.
Converted from JAX to PyTorch bfloat16 with the original Franka training model configuration. Model outputs have 50 steps and 32 coordinates; only the first eight coordinates represent robot actions. Use the bundled assets/franka/norm_stats.json and matching training transforms.
Read log.txt for the output layout, normalization boundary, camera/state inputs, and controller conventions requiring confirmation.
- Task
- Robotics
- Library
- openpi
- Created
- Sep 24, 2026
- Updated
- Sep 24, 2026