MiniLLM – Fundamental ML Practice

A hands-on project built to understand the core mechanics behind how large language models like GPT and LLaMA work. This implementation covers four key techniques used in modern LLMs:

RMSNorm – Efficient layer normalization RoPE – Rotary Positional Embeddings GQA – Grouped Query Attention SwiGLU – Activation function for feed-forward layers

This project served as a deep-dive into the building blocks that power today's state-of-the-art language models.