Yang130/DeeR-VLA

This repository contains the models of the paper DeeR-VLA: Dynamic Inference of Multimodal Large Language Models for Efficient Robot Execution.

• Task D model: DeeR_task_D_D-exit-strategy
• Tasks ABC model: DeeR_task_ABC_D-exit-strategy
• Tasks ABCD model: DeeR_task_ABCD_D-exit-strategy
• 9B Task D model: DeeR_9B_task_D_D-exit-strategy

The models are based on the following open-sourced models:

• OpenFlamingo-3B-vitl-mpt1b-langinstruct
• OpenFlamingo-9B-vitl-mpt7b