A step-by-step Jupyter Notebook demonstrating how to build and train a compact small language model (“SLM”) from scratch using the TinyStories dataset. Covers data preparation, BPE tokenization, ...