ci: Add type checks to experiments tests

[sahusiddharth]

resolves: #5147

• The criteria_evaluator_factory function returns type[ExperimentEvaluator], but MyPy cannot properly infer that the dynamically created class is actually instantiable (it sees it as abstract rather than concrete)
• Used # type: ignore comments specifically for these factory-created evaluator instances to bypass MyPy limitations with dynamic class creation
• Leveraged the create_evaluator decorator's type information to ensure proper typing for evaluator functions throughout the test suite

---

Note

Enables MyPy on experiments tests and updates evaluator tests to use the decorator APIs, adding targeted type ignores for dynamically created evaluators.

• Tests / MyPy:
  • Remove tests/unit/datasets/test_experiments.py from tests/mypy.ini exclude, enabling type checks.
• Experiments tests (tests/unit/datasets/test_experiments.py):
  • LLM evaluator instantiations: add # type: ignore to ConcisenessEvaluator(...) and HelpfulnessEvaluator(...) where dynamic types confuse MyPy.
  • Evaluator decorator tests:
    • Use create_evaluator(kind="CODE") and dict input signatures; assert via .evaluate(...).label.
    • Async variant uses .async_evaluate(...) with label assertions and preserved name/kind metadata.

^{Written by Cursor Bugbot for commit 19fa176. This will update automatically on new commits. Configure here.}

[sahusiddharth]

Hi @axiomofjoy,

When you get a moment, could you please take a quick look at this PR and let me know if there’s anything I should modify? Appreciate your time!