fix: update Dockerfile and requirements-dev.txt for MONAI 1.6 tutorial compatibility

[garciadias]

Summary

• Rebase Docker image from nvcr.io/nvidia/pytorch:24.10-py3 to 25.03-py3 to support RTX 5090 (Blackwell, SM_120/CUDA 12.8); remove the now-obsolete torch.patch ONNX revert which was specific to 24.10
• Pin mlflow<3.0 — mlflow 3.x is broken on Python 3.12 due to a relative import in mlflow.utils.uv_utils (from .. import zipp fails at top-level scope); this caused 4 tutorial notebooks to fail in our CI run
• Pin transformers<5.0 — transformers 5.x references torch.float8_e8m0fnu which does not exist in the nv25.03 build of PyTorch 2.7; this caused the HuggingFace tutorial to fail
• Add aim and lightning>=2.0 as declared dependencies in requirements-dev.txt (were previously undeclared but required by tutorial notebooks)
• Rebuild the NVIDIA pip constraint file to retain numpy==1.26.4 (nv25.03 PyTorch compiled against NumPy 1.x) and add setuptools<71 (newer setuptools dropped pkg_resources needed by legacy setup.py in git-sourced packages like MetricsReloaded and segment-anything)
• Remove python_version <= '3.10' caps from cucim, onnxruntime, and transformers — these restrictions were keeping packages out of the Python 3.12 image unnecessarily
• Install papermill, jupytext, autopep8, autoflake, and ipywidgets directly in the Dockerfile so the tutorial runner is self-contained

Context

These changes were identified by running the full MONAI tutorial test suite in a fresh Docker build against a MONAI 1.6 dev branch and comparing results with a native conda reference run (Eric's run, eccefc57). The rerun with stderr captured (runner_output_our_only.logs) confirmed the specific error for each notebook group.

Type of change

• Bug fix
• Feature
• Hotfix
• Spike / exploration
• Documentation
• Refactor

Test plan

• Rebuild Docker image with docker build -t monai_1_6:latest .
• Re-run bash run_our_only.sh 2>&1 | tee runner_output_v2.logs inside the container
• Verify mlflow notebooks pass (R1: 4 notebooks)
• Verify hugging_face/hugging_face_pipeline_for_monai.ipynb passes (R3)
• Verify experiment_management/spleen_segmentation_aim.ipynb passes (R6)

🤖 Generated with Claude Code

[coderabbitai]

📝 Walkthrough

Walkthrough

Updates the MONAI Docker build environment from PyTorch 24.10 to 25.03, rewrites pip constraints to enforce setuptools<71, bumps Python dependencies including transformers (≥4.41.0,<5.0), mlflow (<3.0), and adds aim and lightning>=2.0. Removes the torch/onnx patch step and makes onnxruntime unconditional across Python versions.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title directly summarizes the main change: updating Docker and requirements for MONAI 1.6 tutorial compatibility.
Description check	✅ Passed	The description covers all required template sections with detailed explanations of changes, rationale, and testing plan.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

⚔️ Resolve merge conflicts

• Resolve merge conflict in branch fix/monai_1_6_release_notebooks

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@Dockerfile`:
- Around line 44-46: The Dockerfile currently re-specifies mlflow and
transformers in the extra pip install line, creating a conflicting second source
of truth; remove the "mlflow<3.0" and "transformers<5.0" entries from the python
-m pip install line that also installs
papermill/jupytext/autopep8/autoflake/ipywidgets so that mlflow and transformers
are only resolved from requirements-dev.txt (which pins >=2.12.2,<3.0 and
>=4.41.0,<5.0 respectively); keep the separate install for the other tools
unchanged.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 1a02d9ae-23fb-489b-b718-130352617025

📥 Commits

Reviewing files that changed from the base of the PR and between eccefc5 and 9a5bfeb.

📒 Files selected for processing (2)

• Dockerfile
• requirements-dev.txt

[coderabbitai]

⚠️ Potential issue | 🟠 Major

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
set -euo pipefail

echo "Dockerfile explicit dependency installs:"
rg -n 'pip install.*(papermill|jupytext|autopep8|autoflake|ipywidgets|mlflow|transformers|aim|lightning)' Dockerfile -C1

echo
echo "requirements-dev.txt dependency declarations:"
rg -n '^(papermill|jupytext|autopep8|autoflake|ipywidgets|mlflow|transformers|aim|lightning)\b' requirements-dev.txt

Repository: Project-MONAI/MONAI

Length of output: 771

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

rg -n '^(papermill|jupytext|autopep8|autoflake|ipywidgets|aim|lightning)\b' requirements-dev.txt

Repository: Project-MONAI/MONAI

Length of output: 216

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

rg -n '^(papermill|jupytext|autopep8|autoflake|ipywidgets|mlflow|transformers)\b' requirements-dev.txt

Repository: Project-MONAI/MONAI

Length of output: 254

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

for p in papermill jupytext autopep8 autoflake ipywidgets mlflow transformers; do
  echo "== $p =="
  rg -n "\b${p}\b" requirements-dev.txt || true
done

Repository: Project-MONAI/MONAI

Length of output: 365

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

nl -ba Dockerfile | sed -n '30,70p'

Repository: Project-MONAI/MONAI

Length of output: 106

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

awk 'NR>=30 && NR<=70 {printf "%4d:%s\n", NR, $0}' Dockerfile

Repository: Project-MONAI/MONAI

Length of output: 2552

---

Unify mlflow/transformers pins in Dockerfile (Line 44-46).

requirements-dev.txt already pins mlflow (>=2.12.2,<3.0) and transformers (>=4.41.0,<5.0), but the Dockerfile adds weaker second-install constraints (mlflow<3.0, transformers<5.0), dropping the lower bounds and creating a second source of truth. Keep the separate installs for papermill/jupytext/autopep8/autoflake/ipywidgets (not declared in requirements-dev.txt), and align/remove the extra mlflow/transformers specifiers to match requirements-dev.txt only.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@Dockerfile` around lines 44 - 46, The Dockerfile currently re-specifies
mlflow and transformers in the extra pip install line, creating a conflicting
second source of truth; remove the "mlflow<3.0" and "transformers<5.0" entries
from the python -m pip install line that also installs
papermill/jupytext/autopep8/autoflake/ipywidgets so that mlflow and transformers
are only resolved from requirements-dev.txt (which pins >=2.12.2,<3.0 and
>=4.41.0,<5.0 respectively); keep the separate install for the other tools
unchanged.