Update README.md
Browse files
README.md
CHANGED
@@ -13,6 +13,23 @@ tags:
|
|
13 |
|
14 |
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
|
15 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
<!--- Describe your model here -->
|
17 |
|
18 |
## Usage (Sentence-Transformers)
|
|
|
13 |
|
14 |
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
|
15 |
|
16 |
+
Some post-training logs:
|
17 |
+
```
|
18 |
+
2024-04-09 15:33:55 - EmbeddingSimilarityEvaluator: Evaluating the model on sts-dev dataset after epoch 0:
|
19 |
+
2024-04-09 15:33:56 - Cosine-Similarity : Pearson: 0.8567 Spearman: 0.8662
|
20 |
+
2024-04-09 15:33:56 - Manhattan-Distance: Pearson: 0.8540 Spearman: 0.8534
|
21 |
+
2024-04-09 15:33:56 - Euclidean-Distance: Pearson: 0.8572 Spearman: 0.8571
|
22 |
+
2024-04-09 15:33:56 - Dot-Product-Similarity: Pearson: 0.6507 Spearman: 0.6496
|
23 |
+
```
|
24 |
+
```
|
25 |
+
2024-04-09 15:33:56 - EmbeddingSimilarityEvaluator: Evaluating the model on sts-test dataset:
|
26 |
+
2024-04-09 15:33:56 - Cosine-Similarity : Pearson: 0.8294 Spearman: 0.8426
|
27 |
+
2024-04-09 15:33:56 - Manhattan-Distance: Pearson: 0.8274 Spearman: 0.8225
|
28 |
+
2024-04-09 15:33:56 - Euclidean-Distance: Pearson: 0.8303 Spearman: 0.8255
|
29 |
+
2024-04-09 15:33:56 - Dot-Product-Similarity: Pearson: 0.5832 Spearman: 0.5722
|
30 |
+
```
|
31 |
+
The `Cosine-Similarity`, `Spearman` scores are most important.
|
32 |
+
|
33 |
<!--- Describe your model here -->
|
34 |
|
35 |
## Usage (Sentence-Transformers)
|