prepare datasets core

scripts/prepare_core_datasets.py

@@ -13,13 +13,13 @@ tokenizer_path = '../tokenizer'
 
 seqs = [
     (0, 1073741824, 1025, 16000),
-    # (1025, 2049, 2049, 8000),
-    # (2049, 4097, 4097, 4000),
-    # (4097, 8193, 8193, 2000),
-    # (8193, 16385, 16385, 1000),
-    # (16385, 32769, 32769, 500),
-    # (32769, 65537, 65537, 250),
-    # (65537, 131073, 131073, 125),
+    (1025, 2049, 2049, 8000),
+    (2049, 4097, 4097, 4000),
+    (4097, 8193, 8193, 2000),
+    (8193, 16385, 16385, 1000),
+    (16385, 32769, 32769, 500),
+    (32769, 65537, 65537, 250),
+    (65537, 131073, 131073, 125),
 ]
 
 #