Remove self-conditioning and late-stage distortion

[vratins]

Strips self-conditioning (sc_* layers, p_self_cond, use_sc) and late-stage path distortion (use_distortion/p_distort/t_distort/sigma_distort) from the flow model, matcher, integrators, and CLIs.X

Summary by CodeRabbit

• Changes
  • Simplified training and inference workflows by removing self-conditioning and distortion options.
  • Updated sampling, integration, and model execution to use streamlined configurations.
• Bug Fixes
  • Improved consistency across training, evaluation, inference, and sampling interfaces.
• Tests
  • Updated automated tests to match the simplified workflows and removed obsolete self-conditioning and distortion scenarios.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: d8f00e61-c56d-4b0a-afce-629668b9ec7d

📥 Commits

Reviewing files that changed from the base of the PR and between 6955e82 and e12c35e.

📒 Files selected for processing (5)

• scripts/inference.py
• scripts/train.py
• src/flow.py
• tests/test_flow.py
• tests/test_forward.py

💤 Files with no reviewable changes (2)

• scripts/train.py
• scripts/inference.py

Included review availability: Your plan includes up to 1 review per rolling hour; 0 remain after this review.

---

📝 Walkthrough

Walkthrough

The change removes self-conditioning and distortion support from FlowWaterGVP, FlowMatcher, training, inference, and related tests. Public method signatures, command-line options, integration calls, and configuration output are simplified.

Changes

Flow API cleanup

Layer / File(s)	Summary
Simplified model and integration paths src/flow.py	Training, validation, Euler integration, RK4 integration, and sampling no longer use self-conditioning or distortion behavior.
Updated training and inference wiring scripts/train.py, scripts/inference.py	Training and inference no longer define or pass removed configuration and method arguments. Runtime logs and saved metrics no longer include self-conditioning settings.
Updated API and behavior tests tests/test_flow.py, tests/test_forward.py	Tests now use the simplified FlowMatcher and FlowWaterGVP interfaces. Self-conditioning and distortion-specific tests were removed.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Merge Risk: ⚪ Minimal · up to e12c3

The PR removes the specified self-conditioning and late-stage distortion paths without any identified current-head correctness or operational risk; no actionable merge-blocking risk remains.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 57.14% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the pull request's main changes: removing self-conditioning and late-stage distortion.
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.

✨ Finishing Touches 💡 2

📝 Generate docstrings 💡

• Create stacked PR
• Commit on current branch

🛠️ Fix failing CI checks 💡

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch dev_remove_sc_distortion

---

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.}