Update olapp.py

olapp.py

@@ -7,7 +7,7 @@ print("Loading model...")
 
 llm = Llama(
     model_path="/home/oluser/olapp/model-q4_K.gguf",
-    n_ctx=4096,
+    n_ctx=1024,
     n_parts=1,
 )
 
@@ -24,9 +24,9 @@ class OlHandler(BaseHTTPRequestHandler):
 
         output = llm(
           q,
-          max_tokens=32, # Generate up to 32 tokens
+          max_tokens=32,
           echo=False
-        )
+        )["choices"][0]["text"]
         
         self.send_response(200)
         self.end_headers()