Update README.md

README.md

@@ -45,3 +45,33 @@ configs:
   - split: korean
     path: data/korean-*
 ---
+
+
+### 🇰🇷🇺🇸🇯🇵🇨🇳🇪🇸 KoCommonGEN v2 Code-switching
+
+This KoCommonGEN v2 Code-switching dataset consists of 99 samples for numerical commonsense reasoning, which were created relying on machine translation.
+
+The dataset can be found on Hugging Face at: [nlpai-lab/ko_commongen_v2_code_switching](https://huggingface.co/datasets/nlpai-lab/ko_commongen_v2_code_switching)
+
+This dataset contains code-switching data for the following languages:
+- Korean (korean)
+- English (english)
+- Japanese (japan)
+- Chinese (china)
+- Spanish (espanol)
+
+(The code-switching data relies on machine translation, which may result in some inaccuracies.)
+
+
+To load the dataset, you can use the following code:
+
+```python
+from datasets import load_dataset
+
+dataset = load_dataset("nlpai-lab/ko_commongen_v2_code_switching")
+
+# To access a specific language dataset:
+korean_data = dataset['korean']
+english_data = dataset['english']
+# ... and so on for other languages
+```