add pre-training script

Former-commit-id: 8ff96509fa
2025-12-14 10:56:56 +08:00 · 2023-05-29 21:37:22 +08:00
parent 0ab4419b86
commit 4c7c96e656
7 changed files with 169 additions and 21 deletions
--- a/data/dataset_info.json
+++ b/data/dataset_info.json
@@ -93,5 +93,15 @@
            "response": "output",
            "history": "history"
        }
+    },
+    "wiki_demo": {
+        "file_name": "wiki_demo.txt",
+        "file_sha1": "b2288edb05b233e5b35250fd4b308a5fa21fa66d",
+        "columns": {
+            "prompt": "instruction",
+            "query": "",
+            "response": "",
+            "history": ""
+        }
    }
 }