Fix verifier kwarg leaking into draft model constructors

[rahul-tuli]

⚠️ DO NOT MERGE before #672. This PR is stacked on #672 (remove eagle converter) and its base is remove-eagle-converter. Merging this first would pull in #672's changes prematurely and/or land against the wrong base. Merge order: #672 first, then this. Once #672 merges, I will rebase this onto main. The diff here is the 3-line loader change only.

Purpose

Fix TypeError: Eagle3DraftModel.__init__() got an unexpected keyword argument 'verifier' raised during draft model construction in scripts/train.py --from-pretrained.

This breaks the EAGLE3 finetuning smoke test:

tests/e2e/smoke/test_finetuning_sanity.py::test_finetuning_weight_sanity
TypeError: Eagle3DraftModel.__init__() got an unexpected keyword argument 'verifier'

build_draft_model passes verifier=args.verifier_name_or_path to from_pretrained (for the external-checkpoint auto-convert path added in #617). SpeculatorModel.from_pretrained had no verifier parameter, so it stayed in **kwargs, was forwarded to HF's from_pretrained, and reached the model constructor cls(config, **kwargs). Draft models take only config and reject it.

Fix

Make verifier an explicit named parameter of SpeculatorModel.from_pretrained so it is consumed at the loader boundary (never forwarded into the constructor), used directly for auto-conversion, and forwarded through the base-class re-dispatch alongside t2d/d2t.

With the attachment-style EagleSpeculator removed in #672, no model constructor takes verifier anymore, so consuming it is sufficient — no per-model special-casing needed.

Tests

• Failing test fixed: tests/e2e/smoke/test_finetuning_sanity.py::test_finetuning_weight_sanity (covered on CI).
• make style / make quality: pass.
• pytest tests/unit/test_model.py tests/unit/convert/test_entrypoints.py: 20 passed.

Checklist

I have filled in:

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan/results, such as providing test command and pasting the results.
• (Optional) The necessary documentation update.
• I (a human) have written or reviewed the code in this pr to the best of my ability.

[coderabbitai]

📝 Walkthrough

Walkthrough

SpeculatorModel.from_pretrained now accepts a verifier argument, forwards verifier values through external checkpoint conversion, and only passes verifier into subclass construction when the target initializer accepts that keyword. A new utility inspects callables for keyword support.

Changes

Verifier routing in loading

Layer / File(s)	Summary
API contract and docs src/speculators/model.py	SpeculatorModel.from_pretrained adds a verifier parameter and documents its behavior during auto-conversion and config-driven loading.
External checkpoint verifier forwarding src/speculators/model.py	The external-checkpoint conversion branch forwards a verifier path or id, and omits the value when the verifier is already a loaded model instance.
Callable kwarg detection and subclass forwarding src/speculators/utils/util.py, src/speculators/model.py	callable_accepts_kwarg checks callable signatures, and from_pretrained only re-injects verifier into subclass kwargs when the selected __init__ accepts it.

Sequence Diagram(s)

sequenceDiagram
  participant U as User
  participant SPF as "SpeculatorModel.from_pretrained"
  participant ECP as "external-checkpoint conversion path"
  participant CAK as "callable_accepts_kwarg"
  participant TSI as "target subclass __init__"
  U->>SPF: verifier
  SPF->>ECP: verifier string/path or None
  SPF->>CAK: inspect TSI for verifier
  CAK-->>SPF: accepts / rejects
  SPF->>TSI: kwargs with verifier when accepted

Loading

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰 I hopped through loading paths so light,
With verifier tucked in just right.
If kwargs fit, I gave a cheer,
If not, I skipped that little peer.
Hop hop—model magic, bright!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main fix: preventing verifier from leaking into draft model constructors.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
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.
Description check	✅ Passed	The description clearly matches the code changes, explaining the verifier loader fix and the new kwarg-handling helper.

✨ 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 fix/from-pretrained-verifier-kwarg-leak

---

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.}