Update app.py

app.py

@@ -16,15 +16,15 @@ model.to('cpu')
 def generate_text(prompt, temperature, top_p):
     prompt_with_eos = " #CONTEXT# " + prompt + " #TOPIC# "  # Add the string "EOS" to the end of the prompt
     input_tokens = tokenizer.encode(prompt_with_eos, return_tensors='pt')
-    
-    if input_tokens.size(1) > 512:
-        return "ERROR, CONTEXT SIZE EXCEEDED"
 
     input_tokens = input_tokens.to('cpu')
-    
+
     generated_text = prompt_with_eos  # Start with the initial prompt plus "EOS"
     prompt_length = len(generated_text)
 
+    if input_tokens.size(1) > 512:
+        generated = "ERROR, CONTEXT SIZE EXCEEDED"
+
     for _ in range(80):  # Adjust the range to control the number of tokens generated
         with torch.no_grad():
             outputs = model(input_tokens)