Update app.py

app.py

@@ -157,6 +157,7 @@ async def generate(request: Request):
     logger.info("➡️ /generate endpoint received a request.") # Log at the very beginning
     data = await request.json()
     prompt = data.get("prompt", "").strip()
+    max_gen_token = data.get("max_tokens", 800).strip()
 
     if not prompt:
         logger.warning("Prompt cannot be empty in /generate request.")
@@ -190,7 +191,7 @@ async def generate(request: Request):
     try:
         response = llm.create_chat_completion(
             messages=messages_for_llm,
-            max_tokens=800,  # Keep response length short for maximum speed
+            max_tokens=max_gen_token,  # Keep response length short for maximum speed
             temperature=0.7, # Adjust temperature for creativity vs. coherence (0.0-1.0)
             stop=["</s>"] # Stop sequence for TinyLlama Chat
         )