use typed quantize config instead of a raw dict

[shengliangxu]

What does this PR do?

But fix:

Use typed QuantizeConfig instead using raw dict for formal typed ModelOpt configs.

The dict typing was accidental.

Summary by CodeRabbit

• Refactor

  • Quantization recipe configuration is now implemented with a strongly-typed, structured schema that enforces type safety and provides enhanced validation with comprehensive error detection capabilities.

• Tests

  • Updated recipe loading tests to correctly validate quantization configurations when recipes are loaded from directories, fully supporting the new structured object-based configuration format.

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 7acba763-eadf-4201-b38d-8cdba8b2a851

📥 Commits

Reviewing files that changed from the base of the PR and between 14b78ae and 7671387.

📒 Files selected for processing (3)

• examples/llm_ptq/hf_ptq.py
• modelopt/recipe/config.py
• tests/unit/recipe/test_loader.py

---

📝 Walkthrough

Walkthrough

The changes introduce type safety to the quantization configuration by replacing an untyped dictionary with a strongly typed QuantizeConfig object in the PTQ recipe configuration. Related code and tests are updated to work with the structured type instead of plain dictionaries.

Changes

Cohort / File(s)	Summary
Config Type Strengthening modelopt/recipe/config.py	The quantize field in ModelOptPTQRecipe changes from dict[str, Any] with default {} to QuantizeConfig with instantiated default QuantizeConfig(). Adds import of QuantizeConfig and removes unused Any import.
Usage Update examples/llm_ptq/hf_ptq.py	Calls .model_dump() on the recipe.quantize object to convert it to a plain dictionary representation when assigning to quant_cfg for downstream quantization logic.
Test Updates tests/unit/recipe/test_loader.py	Adjusts assertions in test_load_recipe_dir to access recipe.quantize properties as object attributes (recipe.quantize.algorithm, recipe.quantize.quant_cfg) rather than dictionary keys.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

🚥 Pre-merge checks | ✅ 3 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: replacing an untyped dict for quantize configuration with a formally typed QuantizeConfig object.
Security Anti-Patterns	✅ Passed	All three modified files contain no security anti-patterns from SECURITY.md including unsafe deserialization, hardcoded remote code execution flags, dynamic code execution, or security bypass comments.

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

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch shengliangx/use-typed-config

---

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

[github-actions]

PR Preview Action v1.8.1
Preview removed because the pull request was closed.
2026-04-14 01:25 UTC

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 77.44%. Comparing base (202c3d3) to head (d5d179a).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1249      +/-   ##
==========================================
+ Coverage   76.90%   77.44%   +0.54%     
==========================================
  Files         350      350              
  Lines       40524    40524              
==========================================
+ Hits        31166    31385     +219     
+ Misses       9358     9139     -219     

Flag	Coverage Δ
examples	43.80% <100.00%> (+1.17%)	⬆️
gpu	57.43% <100.00%> (-0.11%)	⬇️
unit	55.61% <100.00%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[cjluo-nv]

Small, well-scoped bug fix (6 lines changed across 3 files). Changes ModelOptPTQRecipe.quantize from dict[str, Any] to the proper typed QuantizeConfig, fixing an accidental raw-dict typing. The hf_ptq.py example correctly calls .model_dump() to convert back to a dict where downstream code expects dict operations (.append(), copy.deepcopy(), subscript access). Tests are updated to match the new typed access pattern. The change is correct and straightforward.