#transformers

Essential Math & Concepts for LLM Inference

Back of the envelope calculations to estimate model's GPU memory requirements & insights into HW/SW optimizations