book-training. A complete pipeline for training language models to write in a specific author's style using Supervised Fine-Tuning (SFT) with LoRA.