HF: prepend Space frontmatter (CI-generated, not on main)

Promote Stanford 196-class identifier to canonical submission (#33)

Bake metrics narrative into submission notebook (§2.4–§2.6, §3.3, §4.1, §11.2) (#29)

Pre-render Mermaid diagrams + bake preview outputs into submission notebook (#28)

Merge pull request #25 from theDocWho/checkpoint-30-kaggle-api-token-auth

Colab notebooks: use Kaggle KAGGLE_USERNAME/KAGGLE_KEY env vars

Merge pull request #24 from theDocWho/checkpoint-29-colab-and-pathA-notebooks

Add Colab VMMRdb + Path A (CarDD+HITL nc=6) training notebooks

Merge pull request #23 from theDocWho/checkpoint-28-resume-and-gdrive-sync

Add crash-resume + GDrive sync to identifier-continue training

Update dataset symlink path in notebook

Update ROOT path for vmmrdb dataset location

Merge pull request #22 from theDocWho/checkpoint-27-kaggle-nb-fix

Kaggle notebook: fix clone cell (shell-test syntax) + Internet-On note

Merge pull request #21 from theDocWho/checkpoint-26-kaggle-nb-main

Add Kaggle VMMRdb training notebook (full dataset) — re-land onto main

Merge pull request #20 from theDocWho/checkpoint-25-vmmrdb-identifier

Phase 6: VMMRdb identifier extension (196 → up to ~9,170 make/model/year)

Merge pull request #19 from theDocWho/checkpoint-24-deploy-workflow-fix

deploy-hf: try push first, create-Space on failure (stop spurious exit-22)

Merge pull request #18 from theDocWho/checkpoint-23-multi-car

Multi-car mode: detect every vehicle, identify each, group damage per car

Merge pull request #17 from theDocWho/checkpoint-22-no-car-fix

Fix demo error on 'no car' (and 'no image'): return 6 values to match outputs

Merge pull request #16 from theDocWho/checkpoint-21-variant-d-graceful

Variant D: fail gracefully when seg weights are absent (HF Space hotfix)

Merge pull request #15 from theDocWho/checkpoint-20-mask-area-xgboost

XGBoost(C): mask-area cost feature from the YOLOv8-seg model

Merge pull request #14 from theDocWho/checkpoint-19-parts-aware-costing

Variant D: parts-aware costing (YOLOv8-seg damage ∩ parts → catalog)

Merge pull request #13 from theDocWho/checkpoint-16-deployable-gate-identifier

docs: clarify COCO is pretrained-weights-only, not a training dataset

docs: rework architecture explainer for the checkpoint-16 build

Merge pull request #12 from theDocWho/checkpoint-16-deployable-gate-identifier

docs: canvas-driven neural-network architecture explainer

Add make/model confidence floor (0.30) to the deployable Phase 5 stack

Phase 5: activate make/model identification + car-presence gate

Stop ignoring notebooks/from_scratch/ now that it's tracked

ignore notebooks/from_scratch/ — local-only learning notes

Merge pull request #10 from theDocWho/checkpoint-11-production-target-file-fallback

Fix production_target() to fall back to file-on-disk for HF deploy

Merge pull request #9 from theDocWho/checkpoint-10-classifier-with-negatives

ignore local training state files

Bump RELEASE_TAG to v0.2.0 + cache-bust on tag change

Add Stanford Cars as 'no damage' negatives for the classifier

Merge pull request #8 from theDocWho/checkpoint-9-threshold-bugfix-and-honest-feedback

Honor classifier threshold + show explicit no-detection feedback

Merge pull request #7 from theDocWho/checkpoint-8-annotated-output-and-notebooks

Add annotated image output + 7 beginner notebooks

Merge pull request #6 from theDocWho/checkpoint-7-fix-gradio-and-actions-node24