Update README - Run 20251012_060013

README.md

@@ -21,7 +21,7 @@ model-index:
       type: imagenet-1k
     metrics:
     - type: accuracy
-      value: 81.95
+      value: 82.29
 ---
 
 # David: Multi-Scale Crystal Classifier
@@ -51,14 +51,15 @@ as class prototypes with role-weighted similarity computation (Rose Loss).
 ## Performance
 
 ### Best Results
-- **Validation Accuracy**: 81.95%
-- **Best Epoch**: 2
-- **Final Train Accuracy**: 85.03%
+- **Validation Accuracy**: 82.29%
+- **Best Epoch**: 3
+- **Final Train Accuracy**: 86.20%
 
 ### Per-Scale Performance
-- **Scale 256**: 81.95%
-- **Scale 512**: 81.85%
-- **Scale 768**: 81.54%
+- **Scale 256**: 82.29%
+- **Scale 512**: 82.36%
+- **Scale 768**: 82.28%
+- **Scale 1024**: 81.94%
 
 
 ## Usage
@@ -181,4 +182,4 @@ Special thanks to Claude (Anthropic) for debugging assistance.
 
 ---
 
-*Generated on 2025-10-12 06:24:17*
+*Generated on 2025-10-12 06:26:36*