Results, losses & training diagnosis

112347_3a804eb3 · completed_with_weak_training

Scan overlap and projected CAD labels

Automatic weak mappings. Defect names follow CSV row numbers. Poor fits and rejected annotations stay visible; only eligible datasets enter experimental training. Mapping evidence and eligibility JSON · Fit data directory · Partial CAD fits · Not fit data directory

Model checks against weak labels

Training fit and held-out-design agreement are shown separately. These targets are not independently reviewed accuracy. CSV · Evidence JSON

Bold marks the best value in each metric column within its table, including ties. Lower BCE is better; higher AP and IoU are better.

# Training diagnosis: 20261027 / 112347_3a804eb3

State: **completed_with_weak_training**. Trained predictions: 600; scored records: 0; folds with epoch losses: 90.

Reviewed dataset labels: 0.

- cnn2d_unet: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- dgcnn: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- fno2d: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- gaussian_process: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- gaussian_process: requires_training_labels (30 records): GP requires a fitted training posterior; no random/synthetic prior score emitted
- meshgraphnet: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- physics_risk_gnn: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- pointnet: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- sparse_unet: unavailable_backend (30 records): RuntimeError: RuntimeError: Sparse U-Net unavailable: install optional MinkowskiEngine; dense tensors are not a sparse substitute.
- sparse_unet: unavailable_backend (90 records): RuntimeError: Sparse U-Net unavailable: install optional MinkowskiEngine; dense tensors are not a sparse substitute.
- tabular_mlp: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- unet3d: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)
- uvnet: blocked_missing_reviewed_labels (30 records): No eligible CAD-node labels in this split (reviewed, or explicitly enabled weak mappings that passed geometry checks)

1. Review CAD-node risk labels against the exact geometry_sha256 and node order; use each dataset's review_template.json.
2. CSV X/Y/W/H are image pixels. Supply a reviewed pixel-to-CAD mapping before treating them as CAD labels.
3. Run main.py --prepared-manifest <saved prepared_manifest.json> --labels-dir <reviewed-label-folder> to retain the exact labeled geometry.
4. sparse_unet requires MinkowskiEngine. Choose available models with --models if that optional backend is absent.