Fixes and renaming in dps_guidance.py

[CharlelieLrt]

PhysicsNeMo Pull Request

Description

Checklist

• I am familiar with the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.
• The CHANGELOG.md is up to date with these changes.
• An issue is linked to this pull request.
• If I am implementing a new model or modifying any existing model, I have followed the Models Implementation Coding Standards.

Dependencies

Review Process

All PRs are reviewed by the PhysicsNeMo team before merging.

Depending on which files are changed, GitHub may automatically assign a maintainer for review.

We are also testing AI-based code review tools (e.g., Greptile), which may add automated comments with a confidence score.
This score reflects the AI’s assessment of merge readiness and is not a qualitative judgment of your work, nor is
it an indication that the PR will be accepted / rejected.

AI-generated feedback should be reviewed critically for usefulness.
You are not required to respond to every AI comment, but they are intended to help both authors and reviewers.
Please react to Greptile comments with 👍 or 👎 to provide feedback on their accuracy.

[greptile-apps]

Greptile Summary

This PR renames DPSDenoiser → DPSScorePredictor to better reflect the class's role (producing a score-Predictor, not a Denoiser directly) and removes the incorrect Denoiser base-class inheritance. It also updates the math symbol for the guidance strength from \gamma to \rho(t) in the protocol docstring and from \gamma to \Gamma for the SDA scaling parameter in both concrete guidance classes, and adds three inference-mode guard checks that raise a descriptive RuntimeError before attempting requires_grad_().

Key changes:

• DPSDenoiser renamed to DPSScorePredictor across dps_guidance.py and __init__.py — breaking rename with no backward-compatibility alias; callers must update their imports.
• Base class changed from Denoiser to bare object; the class still satisfies the Predictor protocol structurally via @runtime_checkable.
• Added torch.is_inference_mode_enabled() guards in DPSScorePredictor.__call__, ModelConsistencyDPSGuidance.__call__, and DataConsistencyDPSGuidance.__call__.
• The error messages in those guards suggest torch.no_grad() as an alternative fix, but torch.no_grad() does not exit inference mode; only torch.inference_mode(False) does — the message is misleading and should be clarified.

Important Files Changed

Filename	Overview
physicsnemo/diffusion/guidance/dps_guidance.py	Renames DPSDenoiser to DPSScorePredictor (removing the Denoiser base class), updates math symbol gamma→rho/Γ throughout docstrings, and adds inference mode guard checks at three call sites. The guard's error message misleadingly suggests torch.no_grad() as a fix for being inside torch.inference_mode(), which will not actually resolve the issue.
physicsnemo/diffusion/guidance/init.py	Correctly updates the import and all export from DPSDenoiser to DPSScorePredictor in lockstep with the rename in dps_guidance.py.

_{Last reviewed commit: c5bf78a}

[ktangsali]

/blossom-ci

[NickGeneva]

Few comments but approve to unblock