naklecha/llm-inference-optimizations-explained: in this repository, i'm going to implement increasingly complex llm inference optimizations ·
to understand the basics of how llms work, refer to my other respository where i implement llama3 from scratch and explain how llms work one matrix multiplication at a time: https://github.com/naklecha/llama3-from-scratch; this repo of mine has like 15k stars, kinda wild! (this repo is currently wip) note: this is not production quality code, and it will never be, it's just for educational purposes. i like single file, no functions, no classes, codebases that are easy to understand. also, writing code this way is alot more aesthetic. note: i adding c++ variants of these files into the cpp-implementations folder. so far i implemented 4.py in c++. (logs) in this repository, i'm going to implement increasingly complex llm inference optimizations