Update app.py

app.py

@@ -72,8 +72,8 @@ def generate_response(user_message, history):
  #           top_k=50,            # Top-k sampling for efficiency
  #           pad_token_id=veri_tokenizer.eos_token_id,
  #           eos_token_id=veri_tokenizer.eos_token_id,
- #           use_cache=True,      # Enable KV caching for faster generation
- #           repetition_penalty=1.1,  # Reduce repetition
+            use_cache=True,      # Enable KV caching for faster generation
+            repetition_penalty=1.1,  # Reduce repetition
  #           length_penalty=1.0,
  #           early_stopping=True,     # Stop early when appropriate
  #           num_beams=1,            # Greedy search for speed