search

UMass-Foundation-Model / 3D-VLA

[ICML 2024] 3D-VLA: A 3D Vision-Language-Action Generative World Model
https://vis-www.cs.umass.edu/3dvla/
342 stars 13 forks source link

Release code time #1

Open Johnx69 opened 8 months ago

Johnx69 commented 8 months ago

It is a really interesting work!!! When will you release full code for this work?

Thanks!

anyeZHY commented 7 months ago

Hi, thanks for your interest in our work! We are refactoring the code, collecting recently released robotics datasets (such as DROID), and training a larger model. We plan to release our code and data soon. After completing the final training, we will also release our pretrained and finetuned checkpoints.