eniompw/llama-cpp-gpu

Load larger models by offloading model layers to both GPU and CPU

LLaMA.cpp GPU

Offloads some of the model layers to the GPU, allowing larger models to be loaded

• cuBLAS
• Model