Skip to content

running.md

Latest commit

 

History

History
65 lines (45 loc) · 1.3 KB

running.md

File metadata and controls

65 lines (45 loc) · 1.3 KB
Raw

Running the Server

Start the server:

cargo run

Default bind address is 127.0.0.1:3000.

Open the browser UI:

open http://127.0.0.1:3000/

The browser UI template is embedded into the binary.

Podman

Build the container image:

podman build -f Containerfile -t florence2-base-inference-server .

Run it with the downloaded model files mounted at the default path:

podman volume create florence2-data
podman run --rm \
  -p 3000:3000 \
  -v "$PWD/Florence-2-base:/app/Florence-2-base:ro" \
  -v florence2-data:/app/data:U \
  florence2-base-inference-server

The image does not include ONNX model files. Download them separately and mount the Florence-2-base/ directory into /app/Florence-2-base.

To use a custom config file:

podman run --rm \
  -p 3000:3000 \
  -v "$PWD/Florence-2-base:/app/Florence-2-base:ro" \
  -v "$PWD/config.toml:/app/config.toml:ro" \
  -v florence2-data:/app/data:U \
  florence2-base-inference-server

Compose

Build and run with Compose:

podman compose -f compose.yml up --build

Docker Compose works with the same file:

docker compose -f compose.yml up --build

The Compose file mounts ./Florence-2-base into the container read-only and stores runtime data in the florence2-data named volume.