Update README.md

README.md

@@ -18,26 +18,38 @@ This is a fine-tuned **`ResNet-18`** model designed for a 7-class classification
 ---
 
 ## 📈 Evaluation Metrics on Test Data 
-![confusion matrix](images/confusion_matrix.png)
+![confusion matrix](confusion_matrix.png)
 
 Accuracy: 79.92%
+
 Precision: 79.80%
+
 Recall: 79.92%
+
 F1-Score: 79.80%
 
 Classification Report:
+
               precision    recall  f1-score   support
 
            1       0.79      0.81      0.80       329
+           
            2       0.58      0.47      0.52        74
+           
            3       0.51      0.42      0.46       160
+           
            4       0.92      0.90      0.91      1185
+           
            5       0.74      0.78      0.76       478
+           
            6       0.68      0.72      0.70       162
+           
            7       0.75      0.78      0.77       680
-
+           
     accuracy                           0.80      3068
+    
    macro avg       0.71      0.70      0.70      3068
+   
 weighted avg       0.80      0.80      0.80      3068
 
 ## 🧑‍💻 How to Use
@@ -48,7 +60,6 @@ You can load the model weights and architecture for inference or fine-tuning wit
 ```
 
 def get_out_channels(module):
-    """تابعی برای یافتن تعداد کانال‌های خروجی از لایه‌های کانولوشن و BatchNorm"""
     if isinstance(module, nn.Conv2d):
         return module.out_channels
     elif isinstance(module, nn.BatchNorm2d):