Connor Scott
commited on
Commit
•
2ef92a0
1
Parent(s):
ad5d132
Update README.md
Browse files
README.md
CHANGED
@@ -20,19 +20,21 @@ It achieves the following results on the evaluation set:
|
|
20 |
- Loss: 0.1386
|
21 |
- F1: 0.8627
|
22 |
|
23 |
-
|
24 |
|
25 |
-
|
26 |
|
27 |
-
|
|
|
|
|
28 |
|
29 |
-
|
|
|
|
|
30 |
|
31 |
## Training and evaluation data
|
32 |
|
33 |
-
|
34 |
-
|
35 |
-
## Training procedure
|
36 |
|
37 |
### Training hyperparameters
|
38 |
|
|
|
20 |
- Loss: 0.1386
|
21 |
- F1: 0.8627
|
22 |
|
23 |
+
### Intended uses & limitations
|
24 |
|
25 |
+
This model is intended for use in applications requiring named entity recognition in the German language. Example use cases include:
|
26 |
|
27 |
+
- Automated information extraction from German text sources.
|
28 |
+
- Enhancing search engines with entity-aware search capabilities.
|
29 |
+
- Assisting in document organization and tagging in multilingual contexts.
|
30 |
|
31 |
+
**Limitations**:
|
32 |
+
- The model is fine-tuned only on German data from the WikiANN dataset. Its performance on other languages or domains may not be as high.
|
33 |
+
- It may not perform well on informal or highly specialized text without further fine-tuning.
|
34 |
|
35 |
## Training and evaluation data
|
36 |
|
37 |
+
The model was fine-tuned on the German subset of the WikiANN (PAN-X) dataset. This dataset is part of the XTREME benchmark and provides annotated named entities in multiple languages, making it suitable for training cross-lingual NER models.
|
|
|
|
|
38 |
|
39 |
### Training hyperparameters
|
40 |
|