policy number only

data/test.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:482e2265aaca2481e701c35eb30a4238e579df8461406849bc88812d5b3242aa
-size 224354
+oid sha256:6fd3c19c7a55f1ca28f20dec5eb3c9435df4e6bc96182b9fbaca9c8608446dfa
+size 284488

data/train.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12078c13ce335f68bbabe387b5d7769fad4edff8301e34759472c3cc98ef8acc
-size 2014321
+oid sha256:f1c86f7ad020320f143cd944214b72a59e8b88dacb73525e1c5c974fbdf89094
+size 2591030

data_preparation.py

@@ -34,10 +34,10 @@ def save_json(data: dict, file_name: str):
 
 if __name__ == "__main__":
     # split the train data
-    train, test = split_data("policies-qa-training.json", "json")
+    train, test = split_data("train.json", "json")
 
     # save the train split
-    save_json({"data": train}, "train.json")
+    save_json({"data": train}, "data/train.json")
 
     # save the test split
-    save_json({"data": test}, "test.json")
+    save_json({"data": test}, "data/test.json")