How do I troubleshoot and address overfitting issues when fine-tuning a pre-trained transformer model using Hugging Face’s Transformers library? I’ve experimented with various regularization techniques and dataset augmentation methods, but I’m still encountering challenges in achieving a balance between model performance and generalization.