Update README.md

README.md

@@ -28,3 +28,43 @@ configs:
   - split: test
     path: data/test-*
 ---
+
+# Darija-LID
+
+A language identification dataset for Moroccan Darija (ary) and other Arabic dialects.
+
+## Dataset Description
+
+- **Task:** Language Identification (LID)
+- **Language:** Arabic (Moroccan Darija and other dialects)
+- **License:** MIT
+- **Size:** 
+  - Dataset: ~188,805 Moroccan Darija samples, ~77,324 other dialect samples
+  - Train/Test split: 80%/20% 
+
+## Data Fields
+
+- `text`: The input text
+- `dialect`: The dialect code/name (ary for Moroccan Darija, msa for standard Arabic, etc)
+- `label`: Binary classification label (ary/other)
+- `source`: Source dataset identifier
+
+## Data Sources
+
+The dataset is compiled from the following Hugging Face datasets:
+- [atlasia/levantine_dialects](https://huggingface.co/datasets/atlasia/levantine_dialects)
+- [atlasia/Moroccan-Darija-Wiki-Dataset](https://huggingface.co/datasets/atlasia/Moroccan-Darija-Wiki-Dataset)
+- [atlasia/moroccan_darija_domain_classifier_dataset](https://huggingface.co/datasets/atlasia/moroccan_darija_domain_classifier_dataset)
+- [PRAli22/Arabic_dialects_to_MSA](https://huggingface.co/datasets/PRAli22/Arabic_dialects_to_MSA)
+
+
+## Usage
+
+```python
+from datasets import load_dataset
+
+dataset = load_dataset("atlasia/Darija-LID")
+```
+
+## Authors
+Imane Momayiz