This repo reproduces GPT2 by building a trianing pipeline in PyTorch from scratch. Start by following the build_nanogpt lecture by Anjrey Karpathy, changed will be made later to improve functionalities.
This repo reproduces GPT2 by building a trianing pipeline in PyTorch from scratch. Start by following the build_nanogpt lecture by Anjrey Karpathy, changed will be made later to improve functionalities.