train our own model based on our own datasets

[Ericgone]

Hi,
thanks for your excellent work and I have a question. If the pretrained weight doesn't suit my datasets for example I wanna track multiple specific objects which the pretrained model has never seen, what should I do?

[sujanshresstha]

Hello @Ericgone,
You can simply Fine-tune YOLOv9 model on your own dataset and then utilize the resulting weights with this code.
For guidance on training with your dataset, you can access the Colab Notebook available on the YOLOv9 official GitHub page here.