Fix broken image links: use absolute HF URLs and correct learning_curve path

README.md

@@ -199,7 +199,7 @@ We used a **single unified dataset loader** (`UnifiedReceiptDataset`) so all 8 d
 
 ### Loss Curve
 
-![Learning Curve](hub_assets/learning_curve.png)
+![Learning Curve](https://huggingface.co/Awarebeyond/receipt-donut/resolve/main/learning_curve.png)
 
 The model converged around **Epoch 9**. Training was stopped early because:
 - Validation loss plateaued
@@ -222,13 +222,13 @@ The model converged around **Epoch 9**. Training was stopped early because:
 
 Below are real model outputs on the validation set (Original Image vs. Predicted JSON).
 
-![Sample 1](hub_assets/sample_result_0.png)
+![Sample 1](https://huggingface.co/Awarebeyond/receipt-donut/resolve/main/hub_assets/sample_result_0.png)
 *Example 1: Correctly extracted merchant, date, and total.*
 
-![Sample 2](hub_assets/sample_result_1.png)
+![Sample 2](https://huggingface.co/Awarebeyond/receipt-donut/resolve/main/hub_assets/sample_result_1.png)
 *Example 2: Handled a partially blurred receipt with minor date typo.*
 
-![Sample 3](hub_assets/sample_result_2.png)
+![Sample 3](https://huggingface.co/Awarebeyond/receipt-donut/resolve/main/hub_assets/sample_result_2.png)
 *Example 3: Multi-line address and tax amount correctly parsed.*
 
 ---
@@ -256,7 +256,7 @@ Since this is a **generative text model** (not a classifier), a traditional conf
 | `total` | **47.4%** (283/597) | 7.9% (47) | 44.7% (267) | **Hardest field** — model confuses it with subtotal |
 | `address` | **100.0%** (597/597) | 0.0% (0) | 0.0% (0) | **Test set has 0 address labels** — model correctly abstains |
 
-![Field Confusion Matrix](hub_assets/field_confusion_matrix.png)
+![Field Confusion Matrix](https://huggingface.co/Awarebeyond/receipt-donut/resolve/main/hub_assets/field_confusion_matrix.png)
 
 ### Overall Performance