Just a smol model to detect kawaii catto <3
Python >= 3.8
Pytorch >= 1.8
YOLOv5
Using a GPU is not required but is highly recommended.
(Incase you use GPU, get CUDA 11.1 for maximum compatibility with Pytorch 1.8)
Start by cloning YOLOv5 and installing the requirements.
pip install -r https://raw.githubusercontent.com/ultralytics/yolov5/master/requirements.txt
Get the YoloV5s model from here and put it in the same directory.
https://github.com/ultralytics/yolov5/releases
(Any model works, but we recommend using YoloV5s)\
You can get the dataset we labelled here manually, or just let the code do it automatically.
https://github.com/VioletEqz/Catto/releases/download/Dataset/catset.zip
Then run the following to start training.
#python train.py --img 320 --batch 15 --epochs 6 --data catset.yaml --worker 0 --weights yolov5s.pt