Non usable python-gunicorn docker slimed image

Hi all. I have been several days trying to reduce my image size using slim. The image contains three important directories in / that I want to preserve: /code, /venv and /models. It consists of a ubuntu-based cuda image that runs a gunicorn python app that exposes an API.

This is the command that I'm running to slim the image:

slim build \                                                                                                                                                                                                                          (AI-792-investigate-how-we-can-reduce-docker-image-size|✚3…103)
--publish-exposed-ports \
--expose 8080 \
--compose-env-file .env \
--http-probe=false \
--include-path /code \
--include-path /bin \
--include-path /models \
--include-path /venv \
ai-backend:latest-gpu
cmd=build info=exec message='changing continue-after from probe to nothing because http-probe is disabled'
cmd=build info=exec message='changing continue-after to enter'
cmd=build state=started
cmd=build info=params target='ai-backend:latest-gpu' continue.mode='enter' rt.as.user='true' keep.perms='true' tags='' target.type='image'
cmd=build state=image.inspection.start
cmd=build info=image id='sha256:f123cc0e8ad9821f81801019164333e31cc7cbdc8fb170b0e011cfd72d8e6412' size.bytes='14583660396' size.human='15 GB'
cmd=build info=image.stack index='0' name='base-cudnn8-cuda11-ubuntu20.04:98360d7dde8a43a2c4d937cba0573ae7ef5e356b' id='sha256:286383a731a373c0991793a46b9e59cae715e9dd238c5ac26045ce6eb944d4a3'
cmd=build info=image.stack index='1' name='ai-backend:0.14.1-cuda11-98360d7dde8a43a2c4d937cba0573ae7ef5e356b' id='sha256:f123cc0e8ad9821f81801019164333e31cc7cbdc8fb170b0e011cfd72d8e6412'
cmd=build state=image.inspection.done
cmd=build state=container.inspection.start
cmd=build info=container id='bb9a5c5d6fce468e420317dfa5cdd6b2286182d24624ed23352d895e2963857b' status='created' name='slimk_102841_20230912091453'
cmd=build info=container status='running' name='slimk_102841_20230912091453' id='bb9a5c5d6fce468e420317dfa5cdd6b2286182d24624ed23352d895e2963857b'
cmd=build info=container message='obtained IP address' ip='172.17.0.2'
cmd=build info=cmd.startmonitor status='sent'
cmd=build info=event.startmonitor.done status='received'
cmd=build info=container id='bb9a5c5d6fce468e420317dfa5cdd6b2286182d24624ed23352d895e2963857b' target.port.list='8080' target.port.info='8080/tcp => 0.0.0.0:8080' message='YOU CAN USE THESE PORTS TO INTERACT WITH THE CONTAINER' name='slimk_102841_20230912091453'
cmd=build info=continue.after mode='enter' message='provide the expected input to allow the container inspector to continue its execution'
cmd=build prompt='USER INPUT REQUIRED, PRESS <ENTER> WHEN YOU ARE DONE USING THE CONTAINER'

cmd=build state=container.inspection.finishing
cmd=build state=container.inspection.artifact.processing
cmd=build state=container.inspection.done
cmd=build state=building message="building optimized image" engine=internal
cmd=build state=completed
cmd=build info=results status='MINIFIED' by='3.28X' size.original='15 GB' size.optimized='4.5 GB'
cmd=build info=results image.size='4.5 GB' has.data='true' image.name='ai-backend.slim'
cmd=build info=results artifacts.location='/tmp/slim-state/.slim-state/images/f123cc0e8ad9821f81801019164333e31cc7cbdc8fb170b0e011cfd72d8e6412/artifacts'
cmd=build info=results artifacts.report='creport.json'
cmd=build info=results artifacts.dockerfile.reversed='Dockerfile.fat'
cmd=build info=results artifacts.dockerfile.optimized='Dockerfile'
cmd=build info=results artifacts.seccomp='ai-backend-seccomp.json'
cmd=build info=results artifacts.apparmor='ai-backend-apparmor-profile'
cmd=build state=done
cmd=build info=commands message='use the xray command to learn more about the optimize image'
cmd=build info=report file='slim.report.json'
app='slim' message='GitHub Discussions' info='https://github.com/slimtoolkit/slim/discussions'
app='slim' message='Join the CNCF Slack channel to ask questions or to share your feedback' info='https://cloud-native.slack.com/archives/C059QP1RH1S'
app='slim' message='Join the Discord server to ask questions or to share your feedback' info='https://discord.gg/9tDyxYS'
app='slim' message='Join the Gitter channel to ask questions or to share your feedback' info='https://gitter.im/docker-slim/community'

This is the entrypoint defined in the image, defined in /docker-entrypoint.sh:

#! /bin/bash

set -eo pipefail
source /venv/bin/activate

if [ -z "$1" ]; then
echo "Please, add a command to run this docker image"
exit 1
fi

$@

When I run the slim build command I interact with the container to run the following gunicorn command:

/docker-entrypoint.sh "gunicorn --forwarded-allow-ips '''*''' -b 0.0.0.0:8080 --max-requests 5 --threads 1 -k uvicorn.workers.UvicornWorker --preload bin.api:main()"

From outside the container I query the api using curl and it answers correctly. I also run the python interpreter to make slim understand that I want it. Anyway, when I press enter and slim creates the slim image, I try to enter into a container but the image has become unusable:

> docker run -p 8080:8080 -ti ai-backend.slim:latest bash
(venv) /docker-entrypoint.sh "gunicorn --forwarded-allow-ips '''*''' -b 0.0.0.0:8080 --max-requests 5 --threads 1 -k uvicorn.workers.UvicornWorker --preload bin.api:main()"
/docker-entrypoint.sh: /venv/bin/gunicorn: /venv/bin/python3: bad interpreter: No such file or directory

I would be very grateful if someone could help me, please.

Reproduce the slim build with /docker-entrypoint.sh and the shown gunicorn command, then inspect the generated optimized image and the /venv/bin/gunicorn entry point. Compare the working container with the slim image; done means the slim image starts the API successfully and /venv/bin/gunicorn no longer reports a missing Python interpreter.