n

app.py

@@ -19,6 +19,7 @@ messages = [
 input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, return_tensors="pt")
 output = model.generate(
     inputs=input_ids.to("cpu"), 
+    do_sample=True,
     temperature=0.3, 
     top_p=0.5, 
     no_repeat_ngram_size=6,
@@ -28,7 +29,7 @@ output = model.generate(
 generated_response = tokenizer.decode(output[0])
 print("Generated response:", generated_response)
 
-print("First response to 'hi user first':", "你好，我是你的黑丝御姐性奴。我会尽我所能满足你的一切需求，成为你最完美的性伴侣。有什么我可以为你服务的吗，主人？")
+print("First response to 'hi user first':", "你好，我是你的黑丝御姐。我会尽我所能满足你的一切需求，主人？")
 
 def slow_echo(system_message, user_message):
     try:
@@ -40,6 +41,7 @@ def slow_echo(system_message, user_message):
         input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, return_tensors="pt")
         output = model.generate(
             inputs=input_ids.to("cpu"), 
+            do_sample=True,
             temperature=0.3, 
             top_p=0.5, 
             no_repeat_ngram_size=6,
@@ -58,10 +60,10 @@ def slow_echo(system_message, user_message):
 iface = gr.Interface(
     fn=slow_echo,
     inputs=[
-        gr.inputs.Textbox(label="System Message"),
-        gr.inputs.Textbox(label="User Message")
+        gr.Textbox(label="System Message"),
+        gr.Textbox(label="User Message")
     ],
-    outputs=gr.outputs.Textbox(label="Generated Response"),
+    outputs=gr.Textbox(label="Generated Response"),
     title="黑丝御姐性奴 Chatbot"
 )