Use `tuple` instead of `list` in `gen_kwargs` to avoid `IterableDataset.shuffle` error

#3
by rohfle - opened
Files changed (1) hide show
  1. mnist.py +2 -2
mnist.py CHANGED
@@ -87,14 +87,14 @@ class MNIST(datasets.GeneratorBasedBuilder):
87
  datasets.SplitGenerator(
88
  name=datasets.Split.TRAIN,
89
  gen_kwargs={
90
- "filepath": [downloaded_files["train_images"], downloaded_files["train_labels"]],
91
  "split": "train",
92
  },
93
  ),
94
  datasets.SplitGenerator(
95
  name=datasets.Split.TEST,
96
  gen_kwargs={
97
- "filepath": [downloaded_files["test_images"], downloaded_files["test_labels"]],
98
  "split": "test",
99
  },
100
  ),
 
87
  datasets.SplitGenerator(
88
  name=datasets.Split.TRAIN,
89
  gen_kwargs={
90
+ "filepath": (downloaded_files["train_images"], downloaded_files["train_labels"]),
91
  "split": "train",
92
  },
93
  ),
94
  datasets.SplitGenerator(
95
  name=datasets.Split.TEST,
96
  gen_kwargs={
97
+ "filepath": (downloaded_files["test_images"], downloaded_files["test_labels"]),
98
  "split": "test",
99
  },
100
  ),