added name to training sample conversion for helpsteer and oasst

train/prepare/oasst2/oasst22jsonl.py

@@ -1,5 +1,7 @@
 import json
 import os
+import random
+from typing import AnyStr
 
 # parsing OA data files with oasst_data helpers
 from oasst_data import read_message_trees, ExportMessageNode
@@ -10,23 +12,28 @@ this_dir = os.path.dirname(os.path.abspath(__file__))
 input_file_path = this_dir + '/2023-11-05_oasst2_all.trees.jsonl.gz'
 
 role_dict = {'prompter': 'user', 'assistant': 'assistant'}
+user_names = ['Adam', 'Alice', 'Anne', 'Bob', 'Charlie', 'Cody', 'Corinna', 'Cynthia', 'Fred', 'Grace', 'Jane', 'Paul',
+              'Rachel', 'Ramesh']
 conversations = []
 
 
-def visit(node: ExportMessageNode, parents: [ExportMessageNode]):
+def visit(node: ExportMessageNode, parents: [ExportMessageNode], user: AnyStr):
     new_parents = parents + [node]
     if not node.replies:  # end of conversation
-        conversations.append({'messages': [{'role': role_dict[p.role], 'content': p.text} for p in new_parents]})
+        conversations.append({'messages': [{'role': role_dict[p.role],
+                                            'name': user if role_dict[p.role] != 'assistant' else 'assistant',
+                                            'content': p.text
+                                            } for p in new_parents]})
     else:
         for reply in node.replies:
-            visit(reply, new_parents)
+            visit(reply, new_parents, user)
 
 
 for tree in read_message_trees(input_file_path):
     if tree.prompt.lang not in ['en']:  # filtering by language tag (optional)
         continue
 
-    visit(tree.prompt, [])
+    visit(tree.prompt, [], user_names[random.randint(0, len(user_names) - 1)])
 
 print(conversations[0])