jgrosjean-mathesis
/

sentence-swissbert

Sentence Similarity

Inference Endpoints

Model card Files Files and versions Community

jgrosjean commited on Feb 26

Commit

2e83ad6

•

1 Parent(s): e2f0db1

Update README.md

Files changed (1) hide show

README.md +8 -8

README.md CHANGED Viewed

@@ -166,14 +166,14 @@ Sentence SwissBERT achieves comparable or better results as the best-performing
 | Evaluation task        |Swissbert |           |Sentence Swissbert|           |Sentence-BERT|           |
 |------------------------|----------|-----------|------------------|-----------|-------------|-----------|
 |                        |accuracy  |f1-score   |accuracy          |f1-score   |accuracy     |f1-score   |
-| Semantic Similarity DE | 87.20 %  | -         |**93.40 %**       |    -      |  91.80 %    |    -      |
-| Semantic Similarity FR | 84.97 %  | -         |**93.99 %**       |    -      |  93.19 %    |    -      |
-| Semantic Similarity IT | 84.17 %  | -         |**92.18 %**       |    -      |  91.58 %    |    -      |
-| Semantic Similarity RM | 83.17 %  | -         |**91.58 %**       |    -      |  73.35 %    |    -      |
-| Text Classification DE |          |        %  |                  |**78.49 %**|             |  77.23 %  |
-| Text Classification FR |          |        %  |                  |**77.18 %**|             |  76.83 %  |
-| Text Classification IT |          |        %  |                  |  76.65 %  |             |**76.90 %**|
-| Text Classification RM |          |        %  |                  |**77.20 %**|             |  65.35 %  |
 #### Baseline

 | Evaluation task        |Swissbert |           |Sentence Swissbert|           |Sentence-BERT|           |
 |------------------------|----------|-----------|------------------|-----------|-------------|-----------|
 |                        |accuracy  |f1-score   |accuracy          |f1-score   |accuracy     |f1-score   |
+| Semantic Similarity DE | 87.20 %  |    --     |  **93.40 %**     |    --     |  91.80 %    |    --     |
+| Semantic Similarity FR | 84.97 %  |    --     |  **93.99 %**     |    --     |  93.19 %    |    --     |
+| Semantic Similarity IT | 84.17 %  |    --     |  **92.18 %**     |    --     |  91.58 %    |    --     |
+| Semantic Similarity RM | 83.17 %  |    --     |  **91.58 %**     |    --     |  73.35 %    |    --     |
+| Text Classification DE |   --     |           |       --         |**78.49 %**|     --      |  77.23 %  |
+| Text Classification FR |   --     |           |       --         |**77.18 %**|     --      |  76.83 %  |
+| Text Classification IT |   --     |           |       --         |  76.65 %  |     --      |**76.90 %**|
+| Text Classification RM |   --     |           |       --         |**77.20 %**|     --      |  65.35 %  |
 #### Baseline