Update to full V5.9.2 scores (all three rubric families, n=200)

Add evaluator models (9B step100, 27B step296): checkpoints, benchmark, usage

Update bench-200 final full-run scores (n=200 all models) + eval artifacts

Update model card: add 4 data41287 checkpoints, fix dataset revision, add bench-200 scores

Update model card: drop removed 9B ep20 arm

Update model card: drop removed 27B arm, add weight decay column, generalize naming rule

Add model card: naming convention and checkpoint index

initial commit