Upload README.md with huggingface_hub

README.md

@@ -6,21 +6,17 @@ dataset_info:
       sequence: uint8
   - name: label
     dtype: int64
-  - name: embedding32
-    sequence: float32
-  - name: embedding16
-    sequence: float32
-  - name: embedding8
+  - name: embedding
     sequence: float32
   splits:
   - name: train
-    num_bytes: 68640000
+    num_bytes: 62400000
     num_examples: 60000
   - name: test
-    num_bytes: 11440000
+    num_bytes: 10400000
     num_examples: 10000
-  download_size: 33824833
-  dataset_size: 80080000
+  download_size: 29807470
+  dataset_size: 72800000
 ---
 # Dataset Card for "MNIST-preprocessed"