Update app.py

app.py

@@ -23,7 +23,7 @@ from token_bucket import Limiter, MemoryStorage
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
 # Rate limiting configuration
-MAX_MODEL_CALLS_PER_MINUTE = 12  # Conservative buffer below 15 RPM
+MAX_MODEL_CALLS_PER_MINUTE = 10  # Conservative buffer below 15 RPM
 RATE_LIMIT = MAX_MODEL_CALLS_PER_MINUTE
 TOKEN_BUCKET_CAPACITY = RATE_LIMIT
 TOKEN_BUCKET_REFILL_RATE = RATE_LIMIT / 60.0  # Tokens per second
@@ -73,7 +73,7 @@ async def process_question(agent, question_text: str, task_id: str, results_log:
     """Process a single question with global rate limiting."""
     submitted_answer = None
     max_retries = 3
-    retry_delay = 50  # Start with Gemini's recommended delay
+    retry_delay = 60  # Start with Gemini's recommended delay
 
     for attempt in range(max_retries):
         try: