fix: match llama-cpp-python defaults (top_k=40, top_p=0.95, rep_penalty=1.0)

generation_config.json

@@ -6,9 +6,9 @@
     151643
   ],
   "pad_token_id": 151643,
-  "repetition_penalty": 1.1,
+  "repetition_penalty": 1.0,
   "temperature": 0.7,
-  "top_k": 20,
-  "top_p": 0.8,
+  "top_k": 40,
+  "top_p": 0.95,
   "transformers_version": "4.57.6"
 }