Create generation_config.json

Think this file is needed for proper generation and ending the output, I know exl2 relies on it

generation_config.json

@@ -0,0 +1,5 @@
+{
+  "bos_token_id": 128000,
+  "eos_token_id": [128001, 128009],
+  "transformers_version": "4.40.0.dev0"
+}