updating hausa readme
Browse files
README.md
CHANGED
@@ -1,43 +1,45 @@
|
|
1 |
Hugging Face's logo
|
2 |
---
|
3 |
-
language:
|
4 |
datasets:
|
5 |
|
6 |
---
|
7 |
-
# bert-base-multilingual-cased-finetuned-
|
8 |
## Model description
|
9 |
-
**bert-base-multilingual-cased-finetuned-
|
10 |
|
11 |
-
Specifically, this model is a *bert-base-multilingual-cased* model that was fine-tuned on
|
12 |
## Intended uses & limitations
|
13 |
#### How to use
|
14 |
You can use this model with Transformers *pipeline* for masked token prediction.
|
15 |
```python
|
16 |
>>> from transformers import pipeline
|
17 |
-
>>> unmasker = pipeline('fill-mask', model='Davlan/bert-base-multilingual-cased-finetuned-
|
18 |
-
>>> unmasker("
|
19 |
-
|
20 |
-
[{'sequence':
|
21 |
-
'
|
22 |
-
'
|
23 |
-
|
24 |
-
'
|
25 |
-
'
|
26 |
-
|
27 |
-
'
|
28 |
-
'
|
29 |
-
|
30 |
-
'
|
31 |
-
'
|
32 |
-
|
33 |
-
'
|
34 |
-
'
|
|
|
|
|
35 |
|
36 |
```
|
37 |
#### Limitations and bias
|
38 |
This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.
|
39 |
## Training data
|
40 |
-
This model was fine-tuned on
|
41 |
|
42 |
## Training procedure
|
43 |
This model was trained on a single NVIDIA V100 GPU
|
@@ -45,9 +47,9 @@ This model was trained on a single NVIDIA V100 GPU
|
|
45 |
## Eval results on Test set (F-score)
|
46 |
Dataset|F1-score
|
47 |
-|-
|
48 |
-
|
49 |
MasakhaNER |80.82
|
50 |
-
|
51 |
|
52 |
### BibTeX entry and citation info
|
53 |
By David Adelani
|
|
|
1 |
Hugging Face's logo
|
2 |
---
|
3 |
+
language: ha
|
4 |
datasets:
|
5 |
|
6 |
---
|
7 |
+
# bert-base-multilingual-cased-finetuned-hausa
|
8 |
## Model description
|
9 |
+
**bert-base-multilingual-cased-finetuned-hausa** is a **Hausa BERT** model obtained by fine-tuning **bert-base-multilingual-cased** model on Hausa language texts. It provides **better performance** than the multilingual BERT on text classification and named entity recognition datasets.
|
10 |
|
11 |
+
Specifically, this model is a *bert-base-multilingual-cased* model that was fine-tuned on Hausa corpus.
|
12 |
## Intended uses & limitations
|
13 |
#### How to use
|
14 |
You can use this model with Transformers *pipeline* for masked token prediction.
|
15 |
```python
|
16 |
>>> from transformers import pipeline
|
17 |
+
>>> unmasker = pipeline('fill-mask', model='Davlan/bert-base-multilingual-cased-finetuned-hausa')
|
18 |
+
>>> unmasker("Shugaban [MASK] Muhammadu Buhari ya amince da shawarar da ma’aikatar sufuri karkashin jagoranci")
|
19 |
+
|
20 |
+
[{'sequence':
|
21 |
+
'[CLS] Shugaban Nigeria Muhammadu Buhari ya amince da shawarar da ma [UNK] aikatar sufuri karkashin jagoranci [SEP]',
|
22 |
+
'score': 0.9762618541717529,
|
23 |
+
'token': 22045,
|
24 |
+
'token_str': 'Nigeria'},
|
25 |
+
{'sequence': '[CLS] Shugaban Ka Muhammadu Buhari ya amince da shawarar da ma [UNK] aikatar sufuri karkashin jagoranci [SEP]', 'score': 0.007239189930260181,
|
26 |
+
'token': 25444,
|
27 |
+
'token_str': 'Ka'},
|
28 |
+
{'sequence': '[CLS] Shugaban, Muhammadu Buhari ya amince da shawarar da ma [UNK] aikatar sufuri karkashin jagoranci [SEP]', 'score': 0.001990817254409194,
|
29 |
+
'token': 117,
|
30 |
+
'token_str': ','},
|
31 |
+
{'sequence': '[CLS] Shugaban Ghana Muhammadu Buhari ya amince da shawarar da ma [UNK] aikatar sufuri karkashin jagoranci [SEP]', 'score': 0.001566368737258017,
|
32 |
+
'token': 28682,
|
33 |
+
'token_str': 'Ghana'},
|
34 |
+
{'sequence': '[CLS] Shugabanmu Muhammadu Buhari ya amince da shawarar da ma [UNK] aikatar sufuri karkashin jagoranci [SEP]', 'score': 0.0009375187801197171,
|
35 |
+
'token': 11717,
|
36 |
+
'token_str': '##mu'}]
|
37 |
|
38 |
```
|
39 |
#### Limitations and bias
|
40 |
This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.
|
41 |
## Training data
|
42 |
+
This model was fine-tuned on [Hausa CC-100] (http://data.statmt.org/cc-100/)
|
43 |
|
44 |
## Training procedure
|
45 |
This model was trained on a single NVIDIA V100 GPU
|
|
|
47 |
## Eval results on Test set (F-score)
|
48 |
Dataset|F1-score
|
49 |
-|-
|
50 |
+
Hausa GV NER |80.34
|
51 |
MasakhaNER |80.82
|
52 |
+
VOA Hausa Textclass |80.66
|
53 |
|
54 |
### BibTeX entry and citation info
|
55 |
By David Adelani
|