Update app.py

app.py

@@ -469,7 +469,6 @@ def get_answer_from_model(model_id, question, consecutive_failures, failure_thre
     time.sleep(time_sleep) # Small delay
     duration = time.time() - start_time # Calculate duration
     print(f"Answer generation by \"{model_id}\": {duration:.2f}s") # Print answer generation duration separately as requested - as requested
-    st.write(f"Answer generation by \"{model_id}\": {duration:.2f}s") # Print answer generation duration separately as requested - as requested
 
     return answer, duration # Return answer and duration
 
@@ -684,6 +683,7 @@ def run_benchmark(hf_models, topics, difficulties, t, model_config, token=None):
             for future in concurrent.futures.as_completed(answer_futures):
                 model_id = active_models[answer_futures.index(future)] # Get model_id based on future index (order is preserved)
                 answer, duration = future.result() # Get both answer and duration - unpack the returned tuple
+                st.write(f"Answer generation by \"{model_id}\": {duration:.2f}s") # Print answer generation duration separately as requested - as requested
                 answers[model_id] = answer
                 answer_durations[model_id] = duration # Store duration - store the duration in the answer_durations dictionary
 
@@ -957,7 +957,7 @@ if st.sidebar.button("Start Benchmark"):
         # Run the benchmark
         try:
             # Update status
-            status_text.text("Benchmark running...\n\nThis will take a few minutes depending on the number of models and iterations chosen. If you want to follow progress in the question-answer-ranking process per each iteration, check the container log (above, next to the \"running\" button).")
+            status_text.text("Benchmark running...\n\nFor more detailed logs, check the container log (above, next to the \"running\" button).")
             
             # Run benchmark and get results
             results, cumulative_avg_rank, total_successful = run_benchmark(