dataloader

nbnn_language_detection.py

@@ -1,19 +1,19 @@
-from datasets import DatasetBuilder, SplitGenerator, load_dataset
+from datasets import DatasetBuilder, DatasetInfo, SplitGenerator, load_dataset
 from datasets.features import Features, Value
-import json  # Required for json.loads
+import json
 
 class MyDataset(DatasetBuilder):
 
     VERSION = "0.1.0"
 
     def _info(self):
-        return {
-            'features': Features({
+        return DatasetInfo(
+            features=Features({
                 'text': Value('string'),
                 'language': Value('string'),
             }),
-            'supervised_keys': ('text', 'language')
-        }
+            supervised_keys=('text', 'language')
+        )
 
     def _split_generators(self, dl_manager):
         files = {
@@ -34,5 +34,4 @@ class MyDataset(DatasetBuilder):
                 yield id, {
                     'text': data['text'],
                     'language': data['language']
-                }