Adding flair swe ner model
Browse files- README.md +55 -0
- pytorch_model.bin +3 -0
README.md
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- flair
|
4 |
+
- token-classification
|
5 |
+
- sequence-tagger-model
|
6 |
+
language: sv
|
7 |
+
datasets:
|
8 |
+
- SUC 3.0
|
9 |
+
widget:
|
10 |
+
- text: "Hampus Londögård bor i Lund och har levererat denna model idag."
|
11 |
+
---
|
12 |
+
|
13 |
+
## Swedish NER in Flair (SUC 3.0)
|
14 |
+
F1-Score: **85.6** (SUC 3.0)
|
15 |
+
B-PRS, I-PRS, B-ORG, I-ORG, B-TME, I-TME, B-WRK, B-LOC, I-LOC, B-EVN, I-EVN, B-MSR, I-MSR, I-WRK, B-OBJ, I-OBJ
|
16 |
+
Predicts 8 tags:
|
17 |
+
| **tag** | **meaning** |
|
18 |
+
|---------------------------------|-----------|
|
19 |
+
| PRS| cardinal value |
|
20 |
+
| ORG | organisation name|
|
21 |
+
| TME | time unit |
|
22 |
+
| WRK | building name |
|
23 |
+
| LOC | location name |
|
24 |
+
| EVN | event name |
|
25 |
+
| MSR | measurement unit |
|
26 |
+
| OBJ | object |
|
27 |
+
|
28 |
+
Based on [Flair embeddings](https://www.aclweb.org/anthology/C18-1139/) and LSTM-CRF.
|
29 |
+
---
|
30 |
+
### Demo: How to use in Flair
|
31 |
+
Requires: **[Flair](https://github.com/flairNLP/flair/)** (`pip install flair`)
|
32 |
+
```python
|
33 |
+
from flair.data import Sentence
|
34 |
+
from flair.models import SequenceTagger
|
35 |
+
# load tagger
|
36 |
+
tagger = SequenceTagger.load("flair/ner-english-ontonotes-large")
|
37 |
+
# make example sentence
|
38 |
+
sentence = Sentence(<Insert Text>)
|
39 |
+
# predict NER tags
|
40 |
+
tagger.predict(sentence)
|
41 |
+
# print sentence
|
42 |
+
print(sentence)
|
43 |
+
# print predicted NER spans
|
44 |
+
print('The following NER tags are found:')
|
45 |
+
# iterate over entities and print
|
46 |
+
for entity in sentence.get_spans('ner'):
|
47 |
+
print(entity)
|
48 |
+
```
|
49 |
+
This yields the following output:
|
50 |
+
```
|
51 |
+
TODO()
|
52 |
+
```
|
53 |
+
So, the entities "*TODO*" (labeled as a **EVT**), are found in the sentence "TODO".
|
54 |
+
---
|
55 |
+
Please mention londogard if using this models.
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bf20d87ad5710aa5e68a297b0dad9db8f552632e549ed3d025bdbc8600922087
|
3 |
+
size 351555629
|