It is a model made by yolov8 connected with llama which helps it to detect the environment and detect it in real time.
AS the yolo model is remotely installed its lagging a bit so there are two files name: detection.py--> has the llama connected to the yolo model describing the video detection1.py--> here you can see the yolo model detecting the video and with confidence,class.
To run this project, you need to download the YOLO model. You can download it from the official Ultralytics repository:
wget https://github.com/ultralytics/yolov5/releases/download/v6.0/yolov8n.pt
also the project contains my api_key so you will not able to access it remotely
you can create your own going to huggingface models.