FareedKhan-dev/kimi-k3-in-c

Why this repo matters

• •Demonstrates a massive 2.78T parameter LLM running on a single CPU with minimal RAM.
• •Achieves efficient inference without external libraries or GPU, highlighting C99 and SIMD.
• •Offers a portable, low-dependency solution for running large models on constrained hardware.