update trans code

2025-05-09 18:58:19 +08:00
parent a379580943
commit fcf5bd2814
1 changed files with 14 additions and 0 deletions
@@ -0,0 +1,14 @@
+import json
+
+def extract_completions(json_file_path, output_file_path):
+    with open(json_file_path, 'r', encoding='utf-8') as json_file:
+        data = json.load(json_file)
+    
+    completions = [conversion["completion"] for conversion in data]
+
+    with open(output_file_path, 'w', encoding='utf-8') as output_file:
+        for completion in completions:
+            output_file.write(completion + '\n')
+
+extract_completions('corpus/wikipedia-cn-20230720-filtered.json', 'corpus/wikipedia.txt')
+