Update app.py

app.py

@@ -5,18 +5,34 @@ import gradio as gr
 tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-large")
 model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-large")
 
-def predict(input, history=[]):
-    new_user_input_ids = tokenizer.encode(input + tokenizer.eos_token, return_tensors='pt')
-    bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1)
-    history = model.generate(bot_input_ids, max_length=4000, pad_token_id=tokenizer.eos_token_id).tolist()
-    response = tokenizer.decode(history[0]).split("<|endoftext|>")
-    response = [(response[i], response[i+1]) for i in range(0, len(response)-1, 2)]
-    return response, history
-
-gr.Interface(fn=predict,
-             inputs=["text", "state"],
-             outputs=["chatbot", "state"]).launch()
-gr.Interface(..., allow_flagging="never").launch(share=True, inline=True)
-
-iface = gr.Interface(...)  # your interface
-iface.launch(share=True, show_error=True, enable_queue=True, allowed_paths=None, inline=True)
+def predict(user_input, history=[]):
+    # Encode the user input + end-of-text token
+    new_input_ids = tokenizer.encode(user_input + tokenizer.eos_token, return_tensors='pt')
+
+    # Prepare chat history
+    if history:
+        bot_input_ids = torch.cat([torch.tensor(history), new_input_ids], dim=-1)
+    else:
+        bot_input_ids = new_input_ids
+
+    # Generate response
+    chat_history_ids = model.generate(
+        bot_input_ids,
+        max_length=1000,
+        pad_token_id=tokenizer.eos_token_id
+    )
+
+    # Decode the bot's reply
+    response = tokenizer.decode(chat_history_ids[:, bot_input_ids.shape[-1]:][0], skip_special_tokens=True)
+
+    return response, chat_history_ids.tolist()
+
+iface = gr.Interface(
+    fn=predict,
+    inputs=["text", "state"],
+    outputs=["text", "state"],
+    title="DialoGPT Chatbot",
+    description="Chat with DialoGPT-large. Your chat history is preserved.",
+)
+
+iface.launch(share=True, show_error=True, inline=True)