Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
**DBERT** is the first BERT model for the Moroccan Arabic dialect called “Darija”. It is based on the same architecture as BERT-base, but without the Next Sentence Prediction (NSP) objective. This model was trained on a total of ~3 Million sequences of Darija dialect representing 691MB of text or a total of ~100M tokens.
|
2 |
+
|
3 |
+
The model was trained on a dataset issued from three different sources:
|
4 |
+
* Stories written in Darija scrapped from a dedicated website
|
5 |
+
* Youtube comments from 40 different Moroccan channels
|
6 |
+
* Tweets crawled based on a list of Darija keywords.
|
7 |
+
|
8 |
+
More details about DarijaBert are available in the dedicated GitHub repository
|
9 |
+
|
10 |
+
**Loading the model**
|
11 |
+
|
12 |
+
The model can be loaded directly using the Huggingface library:
|
13 |
+
|
14 |
+
```python
|
15 |
+
from transformers import AutoTokenizer, AutoModel
|
16 |
+
DBERT_tokenizer = AutoTokenizer.from_pretrained("Kamel/DBERT")
|
17 |
+
DBERT_Bert_model = AutoModel.from_pretrained("Kamel/DBERT")
|
18 |
+
```
|
19 |
+
|
20 |
+
**Acknowledgments**
|
21 |
+
|
22 |
+
We gratefully acknowledge Google’s TensorFlow Research Cloud (TRC) program for providing us with free Cloud TPUs.
|
23 |
+
|
24 |
+
|