YOLO Model Training

Tutorials:

check the tutorial onhttps://github.com/ultralytics/yolov3/wiki/Train-Custom-Data

Make your own dataset:

label your img on https://www.makesense.ai

your dataset should include the following parts:

data

xx.data
xx.names
xx_test.txt
xx_train.txt

cfg

xx.cfg

img

001
002
...

labels

001
002
...

---

few tips:

cfg and names file and also other file examples can be found on https://github.com/MosHumanoid/YOLO-Model-Training

Training:

environment:

'git clone https://github.com/ultralytics/yolov3.git'

prepare your relies

train:

''' python3 train.py
--data data/xx.data
--cfg cfg/xx.cfg
--weights weights/yolov3-tiny.pt
--epochs 100 '''

yolov3-tiny.pt is a pre-trained model if you don't want it, you can try '--weights " "'