Skip to content

docker-packaging.md

Latest commit

 

History

History
61 lines (40 loc) · 2.02 KB

docker-packaging.md

File metadata and controls

61 lines (40 loc) · 2.02 KB

Docker Packaging Guide

Your container is run once per subject with two volume mounts:

timeout 300 docker run --rm \
  --memory 128g \
  --network none \
  --gpus all \
  -v /path/to/subject/features:/data/features:ro \
  -v /path/to/output:/data/output \
  <your-image>
Mount Mode Contents
/data/features/ read-only NAC-PET, MRI, topogram, metadata for one subject
/data/output/ read-write write your prediction here

Your container must write the predicted CT to /data/output/ct.nii.gz — a NIfTI file in Hounsfield units with the same shape and affine as features/nacpet.nii.gz.

Requirements

  • All model weights baked into the image — network is disabled at runtime (--network none)
  • Output within 5 minutes (hardware: 1× NVIDIA A40, 2× Xeon Gold 6346, 128 GB RAM)
  • No other mounts — do not rely on paths outside /data/

Baseline as a Starting Point

The baseline Dockerfile lives at src/baseline/Dockerfile. It uses a PyTorch base image, installs dependencies from requirements.txt, copies code and weights, and sets the inference script as the entrypoint. You can use it as a template.

FEATURES_DIR = Path("/data/features")
OUTPUT_DIR   = Path("/data/output")

# ...run your model...

# Copy affine from NAC-PET to guarantee shape/affine match
ref = nib.load(str(FEATURES_DIR / "nacpet.nii.gz"))
OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
nib.save(nib.Nifti1Image(pred_hu, ref.affine, ref.header), str(OUTPUT_DIR / "ct.nii.gz"))

Validating Before Submission

Before sending us your image, run it on the 4 validation subjects and upload the predictions to Codabench to confirm your container works end-to-end and your scores look reasonable. We cannot debug containers that fail silently on our infrastructure, so this step is required.

Submitting

See submission-guide.md for full submission instructions, including how to share your image via Docker Hub or a compressed archive.