Fix Docker build: Python 3.12 base image and stale poetry.lock

[sudipdhakal1]

Problem

docker compose up --build (and the manual docker build) currently fail out of the box. Two independent issues:

1. Python version mismatch. The base image tensorflow/tensorflow:2.16.1 ships Python 3.11, but pyproject.toml requires python = "^3.12". poetry install therefore aborts:

The currently activated Python version 3.11.0 is not supported by the project (^3.12).
Poetry was unable to find a compatible version.

2. Stale lockfile. The committed poetry.lock is out of sync with pyproject.toml, so Poetry 2.x refuses to install from it:

pyproject.toml changed significantly since poetry.lock was last generated.
Run `poetry lock` to fix the lock file.

Fix

• Switch the base image to python:3.12-slim-bookworm (satisfying the ^3.12 constraint) and let Poetry install TensorFlow from the lockfile. Add libgomp1, which the TensorFlow CPU wheel needs at runtime.
• Add a poetry lock step before poetry install to refresh the lockfile.

Verification

With these changes the image builds cleanly and the Prediction API serves on :8000:

$ curl 'http://localhost:8000/predict' -X POST -H 'Content-Type: application/json'     --data-raw '{"query":"SELECT * FROM users WHERE id=1 OR 1=1;"}'
{"confidence":0.0019}

Note for maintainers

The Python-version issue could alternatively be fixed by relaxing pyproject.toml to python = "^3.11" and keeping the official tensorflow/tensorflow:2.16.1 base image — a smaller diff. I went with the python:3.12 base image because it keeps the declared ^3.12 requirement intact and is the variant I built and verified end-to-end. Happy to switch to the ^3.11 approach if you prefer to stay on the TensorFlow base image.

🤖 Generated with Claude Code