Validation Accuracy
Loading
Model report
Model Performance
Evaluation results for the QuickSketch CNN, presented as a concise training summary and learning-curve report.
Snapshot
Training Summary
Best Epoch
Loading
Epochs Run
Loading
Early Stopping
Loading
Classes
Loading
Training Samples
Loading
Trainable Parameters
Loading
Values are loaded from the current training artifacts when available.
Learning curves
Training Progress
Loss
Accuracy
Unable to load training artifacts.
Curves are rendered from the saved training history artifact.
Evaluation map
Confusion Matrix
Validation-set class predictions
Confusion matrix image unavailable.
Saved from the held-out validation split for the current released checkpoint.
Class breakdown
Per-Class Metrics
Precision, recall, F1-score, and support
| Class | Precision | Recall | F1-score | Support |
|---|---|---|---|---|
| Loading class metrics... | ||||
Class metrics unavailable.
All values are pulled directly from the saved evaluation metrics artifact.
Model insight
Most Confused Classes
Loading
Confusion insights
Saved misclassification patterns will appear here.
Top off-diagonal confusion pairs from the saved evaluation artifact.