[Unit]
Description=A Llama CPP Server Running GPT OSS 120b

[Container]
# Image is built locally via podman build
Image=localhost/llama-cpp-vulkan:latest

# Downloaded models volume
Volume=/home/ai/models:/models:z

# Ports
PublishPort=8012:8012

# GPU Device
AddDevice=/dev/kfd
AddDevice=/dev/dri

# Server command
Exec=--port 8012 \
    -c 0 \
    --perf \
    --n-gpu-layers all \
    --models-dir /models \
    -m /models/text/qwen2.5-coder-1.5b-instruct/qwen2.5-coder-1.5b-instruct-q8_0.gguf

[Service]
Restart=always
# Extend Timeout to allow time to pull the image
TimeoutStartSec=900

[Install]
# Start by default on boot
WantedBy=multi-user.target default.target