Constrain v7 to prevalidated grounded wording choices

Add deterministic untrusted-choice guard and fresh v6 gate

Replace brittle multi-field output with SAMS v5 wording selector

Rebuild SAMS v4 bounded selector and strict 40-case gate

Resume v3 from preserved adapter and apply 90 percent adapter diagnostic gate

Deploy SAMS v3 targeted quantization-robust retraining

Replace rejected v1 with bounded SAMS v2 training pipeline

Preserve adapter checkpoint and fix GGUF export dependency

Stage validated dataset and base model during free build

Handle blocked optional reference downloads

Upload final artifacts before completing

Add trainer/run_pipeline.py

Add trainer/merge_and_export.py

Add trainer/evaluate.py

Add trainer/train.py

Add trainer/validate_dataset.py

Add trainer/build_dataset.py

Add trainer/download_sources.py

Add trainer/common.py

Add trainer/__init__.py