1.karpathy / llm.cLLM training in simple, raw C/CUDACudaMIT31K total3.8K forks+313 starsfrom 31,074 · score 5.75
2.HigherOrderCO / HVM1HVM1 (2022): a massively parallel, optimal functional runtime in RustCudaMIT11K total437 forks+16 starsfrom 11,335 · score 2.83