How can I fine-tune this model on my own training set?

[Foxerity]

Really amazing work! Could you please consider sharing training code for adapting your project to custom datasets? Thank you.