lostmsu/TurboGPT

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

The simplest, fastest repository for training/finetuning medium-sized GPTs.

LLM training in simple, raw C/CUDA

AI agents running research on single-GPU nanochat training automatically

One-click training of your own GPT. Training a GPT has never been easier for beginners. / 一键预训练+SFT一个属于自己的LLM，0基础训练GPT原来可以这么简单？

A minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training

Zero-dependency C99 GPT-2 engine for edge AI. Sub-1M parameter models train on-device in seconds. Organelle Pipeline Architecture (OPA) coordinates specialised micro-models — 91% win rates on 11 logic games with 30K–160K parameters. Composition beats capacity.

A 49M-parameter Kimi K3-inspired language model trainable on one 8GB GPU

GPT-2-style LLM built from scratch in C/CUDA with hand-written backprop, BPE tokenizer, FlashAttention, pretraining, and SFT.