Update README.md

README.md

@@ -86,16 +86,16 @@ The validation set was used for convergence monitoring and hyperparameter tuning
 * **Regularization:** Dropout (p = 0.5)
 * **Normalization:** Local Response Normalization (LRN)
 
-![Training Loss](./results/loss_curve.png)
 
 ---
 
 ## Performance
 
 * **Test Accuracy:** ~98.42%
+![Training Loss](./results/loss_curve.png)
+![Confusion Matrix](./results/cm.png)
 
 The model performs consistently well across most characters, with minor confusion only between visually similar glyphs.
-![Confusion Matrix](./results/cm.png)
 
 
 ---