File size: 2,956 Bytes
cc503d0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
44
45
46
47
---
language: 
- ar
license: apache-2.0
widget:
 - text: "عامل ايه ؟"
---
# CAMeLBERT-Mix DID NADI Model
## Model description
**CAMeLBERT-Mix DID NADI Model** is a dialect identification (DID) model that was built by fine-tuning the [CAMeLBERT-Mix](https://huggingface.co/CAMeL-Lab/bert-base-arabic-camelbert-mix/) model.
For the fine-tuning, we used the [NADI Coountry-level](https://sites.google.com/view/nadi-shared-task) dataset, which includes 21 labels.
Our fine-tuning procedure and the hyperparameters we used can be found in our paper *"[The Interplay of Variant, Size, and Task Type in Arabic Pre-trained Language Models](https://arxiv.org/abs/2103.06678)."* Our fine-tuning code can be found [here](https://github.com/CAMeL-Lab/CAMeLBERT).

## Intended uses
You can use the CAMeLBERT-Mix DID NADI model as part of the transformers pipeline.
This model will also be available in [CAMeL Tools](https://github.com/CAMeL-Lab/camel_tools) soon.

#### How to use
To use the model with a transformers pipeline:
```python
>>> from transformers import pipeline
>>> did = pipeline('text-classification', model='CAMeL-Lab/bert-base-arabic-camelbert-mix-did-nadi')
>>> sentences = ['عامل ايه ؟', 'شلونك ؟ شخبارك ؟']
>>> did(sentences)
[{'label': 'Egypt', 'score': 0.920274019241333},
 {'label': 'Saudi_Arabia', 'score': 0.26750022172927856}]
```
*Note*: to download our models, you would need `transformers>=3.5.0`.
Otherwise, you could download the models manually.

## Citation
```bibtex
@inproceedings{inoue-etal-2021-interplay,
    title = "The Interplay of Variant, Size, and Task Type in {A}rabic Pre-trained Language Models",
    author = "Inoue, Go  and
      Alhafni, Bashar  and
      Baimukan, Nurpeiis  and
      Bouamor, Houda  and
      Habash, Nizar",
    booktitle = "Proceedings of the Sixth Arabic Natural Language Processing Workshop",
    month = apr,
    year = "2021",
    address = "Kyiv, Ukraine (Online)",
    publisher = "Association for Computational Linguistics",
    abstract = "In this paper, we explore the effects of language variants, data sizes, and fine-tuning task types in Arabic pre-trained language models. To do so, we build three pre-trained language models across three variants of Arabic: Modern Standard Arabic (MSA), dialectal Arabic, and classical Arabic, in addition to a fourth language model which is pre-trained on a mix of the three. We also examine the importance of pre-training data size by building additional models that are pre-trained on a scaled-down set of the MSA variant. We compare our different models to each other, as well as to eight publicly available models by fine-tuning them on five NLP tasks spanning 12 datasets. Our results suggest that the variant proximity of pre-training data to fine-tuning data is more important than the pre-training data size. We exploit this insight in defining an optimized system selection model for the studied tasks.",
}
```