This repository holds the code and report for my final year project at ICL. This project investigates how Large Language Models can be used to decompose high level actions into a set of instructions that can be run on a robot.
This project runs a robot simulated in CoppeliaSim and communicates with it using PyRep. The robot is controlled by a model trained to take in camera input + instruction text and output movement and rotation information. See report for more details and outcome.
Demo:
The repository contains the code to gather training date, train and run the model. The model weights have been omitted due to their large size.