UMass-Foundation-Model / 3D-LLM

Code for 3D-LLM: Injecting the 3D World into Large Language Models
MIT License
952 stars 59 forks source link

Pre-trained weights used to generate the object navigation task #72

Open anindita127 opened 5 months ago

anindita127 commented 5 months ago

Hi, excellent work! I am really interested to know how to generate the object navigation task where the model answers in the form of a position such as "loc_x, loc_y, loc_z". Is the pre-trained model for that provided?