Add train_004 from tar 2

README.md

@@ -22,8 +22,11 @@ dataset_info:
   - name: train_003
     num_bytes: 489097550.0
     num_examples: 5000
-  download_size: 1425581291
-  dataset_size: 1479701915.0
+  - name: train_004
+    num_bytes: 477354902.0
+    num_examples: 5000
+  download_size: 1899071355
+  dataset_size: 1957056817.0
 configs:
 - config_name: default
   data_files:
@@ -31,4 +34,6 @@ configs:
     path: data/train-*
   - split: train_003
     path: data/train_003-*
+  - split: train_004
+    path: data/train_004-*
 ---

data/train_004-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce64ddc1e8317138c99e20a487f3d618f0c01c60a0c5f0527fbdf3eb9d1a04ca
+size 473490064