https://github.com/ngxson/ggml-easy
Thin wrapper around GGML to make life easier
https://github.com/ngxson/ggml-easy
ggml machine-learning
Last synced: 7 months ago
JSON representation
Thin wrapper around GGML to make life easier
- Host: GitHub
- URL: https://github.com/ngxson/ggml-easy
- Owner: ngxson
- License: mit
- Created: 2025-03-15T21:49:57.000Z (7 months ago)
- Default Branch: master
- Last Pushed: 2025-03-15T21:56:42.000Z (7 months ago)
- Last Synced: 2025-03-15T22:27:41.860Z (7 months ago)
- Topics: ggml, machine-learning
- Language: C++
- Homepage:
- Size: 0 Bytes
- Stars: 2
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# ggml-easy
A simple C++ wrapper around [GGML](https://github.com/ggml-org/ggml) to make model loading and execution easier with GPU acceleration support.
## Introduction
`ggml-easy` is a lightweight header-only C++ library that simplifies working with GGML, the tensor library used in projects like llama.cpp. It provides a clean interface for loading GGUF models, creating computation graphs, and executing them on CPU or GPU with minimal boilerplate code.
## Setup
As a header-only library, using ggml-easy is straightforward:
1. Include the headers in your project
2. Make sure you have GGML as a dependency in `CMakeLists.txt`
3. Use the `ggml_easy` namespace in your code
Example:
```cpp
#include "ggml-easy.h"
// Your code here
```
See [demo/basic.cpp](demo/basic.cpp) for a complete example of how to use `ggml-easy` in a project.
## Compile examples
To compile everything inside `demo/*`
```sh
cmake -B build
cmake --build build -j
# output: build/bin/*
```