try to use standard split

README.md

@@ -78,7 +78,7 @@ English
 
 ### Data Splits
 
-trainset, validationset, testset
+train, validation, test
 
 ## Dataset Creation
 

piano_sound_quality.py

@@ -1,6 +1,7 @@
 import io
 import os
 import wave
+import random
 import zipfile
 import datasets
 import requests
@@ -105,19 +106,37 @@ class piano_sound_quality(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
         data_files = dl_manager.download_and_extract(_URLS)
-        split_generator = []
+        dataset = []
 
         for index in _URLS.keys():
-            split_generator.append(
-                datasets.SplitGenerator(
-                    name=index.replace('-', '_'),
-                    gen_kwargs={
-                        "files": dl_manager.iter_files([data_files[index]]),
-                    },
-                )
-            )
-
-        return split_generator
+            for file in dl_manager.iter_files([data_files[index]]):
+                dataset.append(file)
+
+        random.shuffle(dataset)
+        count = len(dataset)
+        p80 = int(0.8 * count)
+        p90 = int(0.9 * count)
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "files": dataset[:p80],
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "files": dataset[p80:p90],
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "files": dataset[p90:],
+                },
+            ),
+        ]
 
     def _generate_examples(self, files):
         for i, path in enumerate(files):