Commit
•
0c4fe75
1
Parent(s):
a9b7cc6
added README
Browse files
README.md
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: sv
|
| 3 |
+
license: mit
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
## Swedish BERT models for sentiment analysis, Sentiment targets.
|
| 7 |
+
[Recorded Future](https://www.recordedfuture.com/) together with [AI Sweden](https://www.ai.se/en) releases a Named Entity Recognition(NER) model for entety detection in Swedish. The model is based on [KB/bert-base-swedish-cased](https://huggingface.co/KB/bert-base-swedish-cased) and finetuned on data collected from various internet sources and forums.
|
| 8 |
+
|
| 9 |
+
The model has been trained on Swedish data and only supports inference of Swedish input texts. The models inference metrics for all non-Swedish inputs are not defined, these inputs are considered as out of domain data.
|
| 10 |
+
|
| 11 |
+
The current models are supported at Transformers version >= 4.3.3 and Torch version 1.8.0, compatibility with older versions are not verified.
|
| 12 |
+
|
| 13 |
+
### Available tags
|
| 14 |
+
|
| 15 |
+
* Location
|
| 16 |
+
* Organization
|
| 17 |
+
* Person
|
| 18 |
+
* Religion
|
| 19 |
+
* Title
|
| 20 |
+
|
| 21 |
+
### Evaluation metrics
|
| 22 |
+
The model had the following metrics when evaluated on test data originating from the same domain as the training data.
|
| 23 |
+
#### F1-score
|
| 24 |
+
| Loc | Org | Per | Nat | Rel | Tit | Total |
|
| 25 |
+
|------|------|------|------|------|------|-------|
|
| 26 |
+
| 0.91 | 0.88 | 0.96 | 0.95 | 0.91 | 0.84 | 0.92 |
|