change dataset to spider and fix output name in multi turn

eosphoros-ai · Jan 1, 2024 · bab9c3b · bab9c3b
1 parent 44d43fd
commit bab9c3b
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 13 deletions.
diff --git a/dbgpt_hub/configs/config.py b/dbgpt_hub/configs/config.py
@@ -47,23 +47,23 @@
 # TODO: BIRD \ WiKiSQL \ ...
 SQL_DATA_INFO = [
     {
-        "data_source": "bird",
-        "train_file": ["train/train.json"],
-        "dev_file": ["dev/dev.json"],
-        "train_tables_file": "train/train_tables.json",
-        "dev_tables_file": "dev/dev_tables.json",
+        "data_source": "spider",
+        "train_file": ["train_spider.json", "train_others.json"],
+        "dev_file": ["dev.json"],
+        "train_tables_file": "tables.json",
+        "dev_tables_file": "tables.json",
         "db_id_name": "db_id",
-        "output_name": "SQL",
+        "output_name": "query",
         "is_multiple_turn": False,
     }
     # {
-    #     "data_source": "spider",
-    #     "train_file": ["train_spider.json", "train_others.json"],
-    #     "dev_file": ["dev.json"],
-    #     "train_tables_file": "tables.json",
-    #     "dev_tables_file": "tables.json",
+    #     "data_source": "bird",
+    #     "train_file": ["train/train.json"],
+    #     "dev_file": ["dev/dev.json"],
+    #     "train_tables_file": "train/train_tables.json",
+    #     "dev_tables_file": "dev/dev_tables.json",
     #     "db_id_name": "db_id",
-    #     "output_name": "query",
+    #     "output_name": "SQL",
     #     "is_multiple_turn": False,
     # }
     # ,

diff --git a/dbgpt_hub/data_process/sql_data_process.py b/dbgpt_hub/data_process/sql_data_process.py
@@ -133,7 +133,7 @@ def decode_json_file(
                         history.append(
                             (
                                 INPUT_PROMPT.format(interaction["utterance"]),
-                                interaction["query"],
+                                interaction[output_name],
                             )
                         )
                 else:  # 单轮