Skip to content

Spec1: matrix harness — {backbone × head} train/eval table #19

Description

@jinmang2

Part of #16 (Spec 1 — Foundation).

Scope

  • One entry (scripts/run_matrix.py or run.py extension): (backbone, head, dataset) → train + eval → log test ROC-AUC + artifacts.
  • Auto-select only compatible pairs (from the text-align guard) → emit {backbone × head} table.

Done when

  • Running the harness produces a populated comparison table for ≥2 backbones × ≥2 heads.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions