Update app.py

app.py

@@ -17,7 +17,7 @@ GREETING = (
 # Constants
 EMBEDDING_MODEL_NAME = "all-MiniLM-L12-v2"
 LLM_MODEL_NAME = "Qwen/Qwen2.5-0.5B-Instruct"
-PUBLICATIONS_TO_RETRIEVE = 3
+PUBLICATIONS_TO_RETRIEVE = 10
 
 
 def embedding(
@@ -77,7 +77,7 @@ def reply(message: str, history: list[str]) -> str:
     # Preprocess message
 
     pipe = transformers.pipeline(
-        "text2text-generation", model="Qwen/Qwen2.5-7B-Instruct", device="cuda"
+        "text2text-generation", model="Qwen/Qwen2.5-7B-Instruct", device="cuda", return_full_text=False
     )
 
     message = preprocess(message, PUBLICATIONS_TO_RETRIEVE)