enjector/microgpt-c

The most atomic way to train and inference a GPT in pure, dependency-free C

Train a tiny GPT in under a minute (CUDA only)

lightweight, standalone C++ inference engine for Google's Gemma models.

Run a 1-billion parameter LLM on a $10 board with 256MB RAM

Local AI API Platform

Pure C++ implementation of several models for real-time chatting on your computer (CPU & GPU)

MiniCPM5: SOTA on-device LLMs, small yet powerful.

High-speed Large Language Model Serving for Local Deployment

CrowdGPT is an opensource AI that aims to be trained without any datacenters