Update README.md

README.md

@@ -20,7 +20,7 @@ dataset_info:
   dataset_size: 91949996.30556343
 ---
 This dataset contains a combination of the following QA datasets:
-Removed all queries with tokenized length > 128 as well as all queries that had one tokenized label > 64. tokenized with gamma-2b tokenizer.
+
 - nq open
 - msmarco 2.1 (first 100k)
 - adverserial qa
@@ -31,4 +31,6 @@ Removed all queries with tokenized length > 128 as well as all queries that had
 - triviaqa
 - wikiqa 
 - freebase_qa
-- squad_v1.1
+- squad_v1.1
+
+Removed all queries with tokenized length > 128 as well as all queries that had one tokenized label > 64. tokenized with gamma-2b tokenizer.