Convert dataset to Parquet

Convert dataset to Parquet.

This dataset uses `tasks`, which are deprecated and will raise an error after the next release of `datasets`. See: https://github.com/huggingface/datasets/pull/6999

README.md

@@ -20,6 +20,7 @@ task_ids:
 paperswithcode_id: fashion-mnist
 pretty_name: FashionMNIST
 dataset_info:
+  config_name: fashion_mnist
   features:
   - name: image
     dtype: image
@@ -37,16 +38,23 @@ dataset_info:
           '7': Sneaker
           '8': Bag
           '9': Ankle boot
-  config_name: fashion_mnist
   splits:
   - name: train
-    num_bytes: 31296655
+    num_bytes: 31049107.0
     num_examples: 60000
   - name: test
-    num_bytes: 5233818
+    num_bytes: 5192560.0
     num_examples: 10000
-  download_size: 30878645
-  dataset_size: 36530473
+  download_size: 36106894
+  dataset_size: 36241667.0
+configs:
+- config_name: fashion_mnist
+  data_files:
+  - split: train
+    path: fashion_mnist/train-*
+  - split: test
+    path: fashion_mnist/test-*
+  default: true
 ---
 
 # Dataset Card for FashionMNIST

fashion_mnist/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb36d8b5b74a491bce99357dcee62a32d347a356ea2a35d1c9057e7a7435ca5a
+size 5175617

fashion_mnist/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5769a70af055b15ada8bd3dad9057557554fdd30e17cba45e7ef0f8bf79c5bb
+size 30931277