21:24
2026-06-28
runtimewire.com
large-language-models
Vincenzo's NanoEuler rebuilds a GPT-2-scale training stack in C and CUDA
Developer Vincenzo released NanoEuler, a from-scratch GPT-2-scale language model training stack written in C and CUDA without PyTorch or autograd, on June 28, 2026. The 116-million-parameter project i…