start.sh

#!/bin/bash

# 設定工作目錄
export WORK="/home/user/app"
cd "$WORK" || exit 1

# 建立 Python 虛擬環境
python3 -m venv venv

# 啟動虛擬環境
source venv/bin/activate

echo "Unzip and download model..."
unzip llama_cpp_avx512.zip > /dev/null 2>&1
wget -O model.gguf https://huggingface.co/Intel/gpt-oss-20b-gguf-q2ks-AutoRound/resolve/main/gpt-oss-20b-32x2.4B-Q2_K_S.gguf > /dev/null 2>&1

echo "Start llama.cpp server..."
./llama-server -m model.gguf \
                --port 8000 \
                --host 0.0.0.0 \
                --threads 2 \
                --ctx-size 4096 \
                --mlock \
                --jinja \
                --temp 0.2 \
                --top-p 0.85 &

# 安裝 Python 套件到虛擬環境
pip install --upgrade pip
pip install gradio openai

echo "Start app.py..."
python app.py

#python3 -m http.server 7860 --bind 0.0.0.0