How do you handle imbalanced datasets when training or fine-tuning generative models especially with class distribution biases

Question

Can you tell me how can i handle imbalanced datasets when training or fine-tuning generative models use python programming to show?