Update README.md

README.md

@@ -59,7 +59,28 @@ It achieves the following results on the evaluation set:
 - Precision: 0.8584
 - Recall: 0.8584
 - F1: 0.8584
-- Accuracy: 0.8584
+- Tokens Accuracy: 0.8584
+- Line Accuracy: 0.9197
+
+## Accuracy at line level
+
+- Line Accuracy: 91.97%
+- Accuracy by label
+  - Caption: 79.42%
+  - Footnote: 68.21%
+  - Formula: 98.02%
+  - List-item: 82.72%
+  - Page-footer: 99.17%
+  - Page-header: 84.18%
+  - Picture: 83.2%
+  - Section-header: 76.92%
+  - Table: 97.65%
+  - Text: 91.17%
+  - Title: 77.46%
+
+![Lines labels vs accuracy (%) of the dataset DocLayNet base of test (model: LiLT base finetuned on DocLayNet base))](https://huggingface.co/pierreguillou/lilt-xlm-roberta-base-finetuned-with-DocLayNet-base-at-linelevel-ml384/resolve/main/docs/lines_labels_accuracy_DocLayNet_base_test_LiLT_base_line_level_384.png)
+
+![Confusion matrix of the labeled lines of the dataset DocLayNet base of test (model: LiLT base finetuned on DocLayNet base)](https://huggingface.co/pierreguillou/lilt-xlm-roberta-base-finetuned-with-DocLayNet-base-at-linelevel-ml384/resolve/main/docs/confusion_matrix_labeled_lines_DocLayNet_base_test_LiLT_base_line_level_384.png)
 
 ## References