Update README.md

README.md

@@ -7,13 +7,13 @@ dataset_info:
     dtype: string
   splits:
   - name: train
-    num_bytes: 1947171484.0
+    num_bytes: 1947171484
     num_examples: 2198247
   - name: validation
-    num_bytes: 24180392.0
+    num_bytes: 24180392
     num_examples: 41623
   download_size: 1035549086
-  dataset_size: 1971351876.0
+  dataset_size: 1971351876
 configs:
 - config_name: default
   data_files:
@@ -21,4 +21,7 @@ configs:
     path: data/train-*
   - split: validation
     path: data/validation-*
+task_categories:
+- text-generation
 ---
+Stories data for the paper [[2505.05755] Insertion Language Models: Sequence Generation with Arbitrary-Position Insertions](https://arxiv.org/abs/2505.05755).