Update README - Run 20251012_065325

README.md

@@ -21,7 +21,7 @@ model-index:
       type: imagenet-1k
     metrics:
     - type: accuracy
-      value: 83.09
+      value: 83.10
 ---
 
 # David: Multi-Scale Crystal Classifier
@@ -51,18 +51,20 @@ as class prototypes with role-weighted similarity computation (Rose Loss).
 ## Performance
 
 ### Best Results
-- **Validation Accuracy**: 83.09%
-- **Best Epoch**: 6
-- **Final Train Accuracy**: 92.04%
+- **Validation Accuracy**: 83.10%
+- **Best Epoch**: 8
+- **Final Train Accuracy**: 94.02%
 
 ### Per-Scale Performance
-- **Scale 256**: 83.09%
-- **Scale 512**: 83.16%
-- **Scale 768**: 83.14%
-- **Scale 1024**: 83.02%
-- **Scale 1280**: 82.94%
-- **Scale 1536**: 82.76%
-- **Scale 1792**: 82.60%
+- **Scale 256**: 83.10%
+- **Scale 512**: 83.24%
+- **Scale 768**: 83.06%
+- **Scale 1024**: 83.03%
+- **Scale 1280**: 82.92%
+- **Scale 1536**: 82.95%
+- **Scale 1792**: 82.98%
+- **Scale 2048**: 82.97%
+- **Scale 2304**: 82.89%
 
 
 ## Usage
@@ -196,4 +198,4 @@ Special thanks to Claude (Anthropic) for debugging assistance.
 
 ---
 
-*Generated on 2025-10-12 07:53:44*
+*Generated on 2025-10-12 07:58:51*