Add README.md

# llama.cpp CUDA Build for Epimetheus (T4 GPU)

This repository provides a Makefile to clone and build [`llama.cpp`](https://github.com/ggerganov/llama.cpp) with CUDA backend support specifically configured for NVIDIA T4 GPUs on Epimetheus servers.

## Prerequisites

Ensure the following are available:

- CUDA 12.9 installed at `/usr/local/cuda-12.9`

- `git`, `cmake`, `make`, and `nvcc` available in `PATH`

- Access to the `git@github.com:ggml-org/llama.cpp.git` repository via SSH

## Build Instructions

For specific version use, for example:

```bash

VERSION=b7650 make install

```