Installation

Step 1: Clone the repository

git clone https://github.com/Yinsongxu/LLM2Jev.git
cd LLM2Jev

Step 2: Choose an installation method

Using uv (recommended)

Choose one backend to install.

SGLang backend (recommended on Linux with a supported NVIDIA GPU):

uv sync --extra sglang

Transformers backend:

uv sync --extra transformers

MLX backend (Apple Silicon):

uv sync --extra mlx

Using pip

Create and activate a virtual environment first:

python -m venv .venv
source .venv/bin/activate

Then install one backend as needed:

python -m pip install -e ".[sglang]"
# Transformers backend:
python -m pip install -e ".[transformers]"
# MLX backend on Apple Silicon:
python -m pip install -e ".[mlx]"

After installation, see the Usage guide to run Python examples or start the SGLang, MLX, or Transformers HTTP service.