File size: 1,183 Bytes
d4ffbcb
3e71317
 
 
d4ffbcb
3e71317
2d39c5c
3e71317
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
---
language: tr
datasets:
- SUNLP-NER-Twitter
---

# bert-loodos-sunlp-ner-turkish

## Introduction
[bert-loodos-sunlp-ner-turkish] is a NER model that was fine-tuned from the loodos/bert-base-turkish-cased model on the SUNLP-NER-Twitter dataset. 

## Training data
The model was trained on the SUNLP-NER-Twitter dataset (5000 tweets). The dataset can be found at https://github.com/SU-NLP/SUNLP-Twitter-NER-Dataset
Named entity types are as follows:
Person, Location, Organization, Time, Money, Product, TV-Show


## How to use bert-loodos-sunlp-ner-turkish with HuggingFace

```python
from transformers import AutoTokenizer, AutoModelForTokenClassification

tokenizer = AutoTokenizer.from_pretrained("busecarik/bert-loodos-sunlp-ner-turkish")
model = AutoModelForTokenClassification.from_pretrained("busecarik/bert-loodos-sunlp-ner-turkish")
```

## Model performances on SUNLP-NER-Twitter test set (metric: seqeval)
Precision|Recall|F1
-|-|-
83.46|85.65|84.53

Classification Report

Entity|Precision|Recall|F1
-|-|-|-
LOCATION|0.82|0.71|0.76
MONEY|0.92|0.76|0.83
ORGANIZATION|0.82|0.87|0.85
PERSON|0.91|0.91|0.91
PRODUCT|0.57|0.33|0.42
TIME|0.86|0.83|0.85
TVSHOW|0.65|0.63|0.64