[FEATURE] Add Regression Tests for Lexical Baseline Models

[federetyk]

Problem

Is your proposal tackling an existing problem or limitation?

• No, it's an addition

Proposal

Add regression tests for the lexical baseline models (BM25Model, TfIdfModel, EditDistanceModel, and RandomRankingModel) that run each model on a small, known benchmark task, and then assert that the resulting metrics match expected values (perhaps within a tolerance range).

The non-deterministic baseline, RandomRankingModel, should use a fixed seed to ensure reproducibility.

Proposal Characteristics

• Type:

  • New Ontology (data source for multiple tasks)
  • New Task(s)
  • New Model(s)
  • New Metric(s)
  • Other

• Area(s) of code: paths, modules, or APIs you expect to touch
  tests/

Additional Context

The current unit tests in #36 verify that the models produce outputs with correct shapes and types. The proposed regression tests would complement these by verifying that the results are correct on a known evaluation scenario. This issue was identified during the review of #36 by @Mattdl.

Implementation

• I plan to implement this in a PR
• I am proposing the idea and would like someone else to pick it up