readme: fix table
Browse files
README.md
CHANGED
|
@@ -59,9 +59,9 @@ Evaluation is performed with SpanMarkers internal evaluation code that uses `seq
|
|
| 59 |
We fine-tune 5 models and upload the model with best F1-Score on development set. Results on development set are
|
| 60 |
in brackets:
|
| 61 |
|
| 62 |
-
| Model
|
| 63 |
-
|
|
| 64 |
-
| GWLMS TEAMS
|
| 65 |
|
| 66 |
The best model achieves a final test score of 87.81%.
|
| 67 |
|
|
|
|
| 59 |
We fine-tune 5 models and upload the model with best F1-Score on development set. Results on development set are
|
| 60 |
in brackets:
|
| 61 |
|
| 62 |
+
| Model | Run 1 | Run 2 | Run 3 | Run 4 | Run 5 | Avg.
|
| 63 |
+
| ----------- | --------------- | --------------- | --------------- | ------------------- | ----------------| ---------------
|
| 64 |
+
| GWLMS TEAMS | (88.76) / 87.85 | (88.54) / 87.77 | (88.41) / 87.98 | (**88.86**) / 87.81 | (88.83) / 88.50 | (88.68) / 87.98
|
| 65 |
|
| 66 |
The best model achieves a final test score of 87.81%.
|
| 67 |
|