gave server more time to process

app.py

@@ -34,9 +34,16 @@ def chat_with_model(model_choice, user_message, chat_history, file=None):
     # Tokenize Input
     inputs = tokenizer(user_message, return_tensors="pt", padding=True, truncation=True, max_length=512)
     
-    # Adjust max_length for longer responses if "PDF Summarizer (T5)" is selected
+    # Adjust max_length and parameters for the PDF summarizer
     max_length = 512 if model_choice == "PDF Summarizer (T5)" else 150
-    outputs = model.generate(**inputs, max_length=max_length, num_beams=5, early_stopping=True)
+    num_beams = 4 if model_choice == "PDF Summarizer (T5)" else 5
+    outputs = model.generate(
+        **inputs, 
+        max_length=max_length, 
+        num_beams=num_beams, 
+        early_stopping=True, 
+        no_repeat_ngram_size=2  # Allow some flexibility in repetition
+    )
     
     response = tokenizer.decode(outputs[0], skip_special_tokens=True)
 
@@ -105,4 +112,4 @@ def create_chat_interface():
 
 if __name__ == "__main__":
     interface = create_chat_interface()
-    interface.launch()
+    interface.launch(server_name="0.0.0.0", server_port=7860, server_timeout=300)