Update README.md

README.md

@@ -1233,8 +1233,8 @@ These datasets can be used to train multilingual sentence embedding models. For
 * Examples:
     ```python
     {
-      "english": "",
-      "non_english": "",
+      "english": "He and his mother will be standing vigil there.”",
+      "non_english": "Él y su madre estarán de vigilia allí”.",
     }
     ```
 * Collection strategy: Processing the data from [yhavinga/ccmatrix](https://huggingface.co/datasets/yhavinga/ccmatrix) and reformatting it in Parquet and with "english" and "non_english" columns.